Guo-Jun Qi

Ziteng Cui, Guo-Jun Qi, Lin Gu et al.

Dark environment becomes a challenge for computer vision algorithms owing to insufficient photons and undesirable noise. To enhance object detection in a dark environment, we propose a novel multitask auto encoding transformation (MAET) model which is able to explore the intrinsic pattern behind illumination translation. In a self-supervision manner, the MAET learns the intrinsic visual structure by encoding and decoding the realistic illumination-degrading transformation considering the physical noise model and image signal processing (ISP). Based on this representation, we achieve the object detection task by decoding the bounding box coordinates and classes. To avoid the over-entanglement of two tasks, our MAET disentangles the object and degrading features by imposing an orthogonal tangent regularity. This forms a parametric manifold along which multitask predictions can be geometrically formulated by maximizing the orthogonality between the tangents along the outputs of respective tasks. Our framework can be implemented based on the mainstream object detection architecture and directly trained end-to-end using normal target detection datasets, such as VOC and COCO. We have achieved the state-of-the-art performance using synthetic and real-world datasets. Code is available at https://github.com/cuiziteng/MAET.

Guangwei Gao, Zixiang Xu, Juncheng Li et al.

Recently, deep convolution neural networks (CNNs) steered face super-resolution methods have achieved great progress in restoring degraded facial details by jointly training with facial priors. However, these methods have some obvious limitations. On the one hand, multi-task joint learning requires additional marking on the dataset, and the introduced prior network will significantly increase the computational cost of the model. On the other hand, the limited receptive field of CNN will reduce the fidelity and naturalness of the reconstructed facial images, resulting in suboptimal reconstructed images. In this work, we propose an efficient CNN-Transformer Cooperation Network (CTCNet) for face super-resolution tasks, which uses the multi-scale connected encoder-decoder architecture as the backbone. Specifically, we first devise a novel Local-Global Feature Cooperation Module (LGCM), which is composed of a Facial Structure Attention Unit (FSAU) and a Transformer block, to promote the consistency of local facial detail and global facial structure restoration simultaneously. Then, we design an efficient Feature Refinement Module (FRM) to enhance the encoded features. Finally, to further improve the restoration of fine facial details, we present a Multi-scale Feature Fusion Unit (MFFU) to adaptively fuse the features from different stages in the encoder procedure. Extensive evaluations on various datasets have assessed that the proposed CTCNet can outperform other state-of-the-art methods significantly. Source code will be available at https://github.com/IVIPLab/CTCNet.

Wenjie Li, Juncheng Li, Guangwei Gao et al.

Lightweight image super-resolution aims to reconstruct high-resolution images from low-resolution images using low computational costs. However, existing methods result in the loss of middle-layer features due to activation functions. To minimize the impact of intermediate feature loss on reconstruction quality, we propose a Feature Interaction Weighted Hybrid Network (FIWHN), which comprises a series of Wide-residual Distillation Interaction Block (WDIB) as the backbone. Every third WDIB forms a Feature Shuffle Weighted Group (FSWG) by applying mutual information shuffle and fusion. Moreover, to mitigate the negative effects of intermediate feature loss, we introduce Wide Residual Weighting units within WDIB. These units effectively fuse features of varying levels of detail through a Wide-residual Distillation Connection (WRDC) and a Self-Calibrating Fusion (SCF). To compensate for global feature deficiencies, we incorporate a Transformer and explore a novel architecture to combine CNN and Transformer. We show that our FIWHN achieves a favorable balance between performance and efficiency through extensive experiments on low-level and high-level tasks. Codes will be available at \url{https://github.com/IVIPLab/FIWHN}.

Wenjie Li, Juncheng Li, Guangwei Gao et al.

Recently, Transformer-based methods have shown impressive performance in single image super-resolution (SISR) tasks due to the ability of global feature extraction. However, the capabilities of Transformers that need to incorporate contextual information to extract features dynamically are neglected. To address this issue, we propose a lightweight Cross-receptive Focused Inference Network (CFIN) that consists of a cascade of CT Blocks mixed with CNN and Transformer. Specifically, in the CT block, we first propose a CNN-based Cross-Scale Information Aggregation Module (CIAM) to enable the model to better focus on potentially helpful information to improve the efficiency of the Transformer phase. Then, we design a novel Cross-receptive Field Guided Transformer (CFGT) to enable the selection of contextual information required for reconstruction by using a modulated convolutional kernel that understands the current semantic information and exploits the information interaction within different self-attention. Extensive experiments have shown that our proposed CFIN can effectively reconstruct images using contextual information, and it can strike a good balance between computational cost and model performance as an efficient model. Source codes will be available at https://github.com/IVIPLab/CFIN.

Ziteng Cui, Yingying Zhu, Lin Gu et al.

Image restoration algorithms such as super resolution (SR) are indispensable pre-processing modules for object detection in low quality images. Most of these algorithms assume the degradation is fixed and known a priori. However, in practical, either the real degradation or optimal up-sampling ratio rate is unknown or differs from assumption, leading to a deteriorating performance for both the pre-processing module and the consequent high-level task such as object detection. Here, we propose a novel self-supervised framework to detect objects in degraded low resolution images. We utilizes the downsampling degradation as a kind of transformation for self-supervised signals to explore the equivariant representation against various resolutions and other degradation conditions. The Auto Encoding Resolution in Self-supervision (AERIS) framework could further take the advantage of advanced SR architectures with an arbitrary resolution restoring decoder to reconstruct the original correspondence from the degraded input image. Both the representation learning and object detection are optimized jointly in an end-to-end training fashion. The generic AERIS framework could be implemented on various mainstream object detection architectures with different backbones. The extensive experiments show that our methods has achieved superior performance compared with existing methods when facing variant degradation situations. Code would be released at https://github.com/cuiziteng/ECCV_AERIS.

Xiao Wang, Yuhang Huang, Dan Zeng et al.

As a representative self-supervised method, contrastive learning has achieved great successes in unsupervised training of representations. It trains an encoder by distinguishing positive samples from negative ones given query anchors. These positive and negative samples play critical roles in defining the objective to learn the discriminative encoder, avoiding it from learning trivial features. While existing methods heuristically choose these samples, we present a principled method where both positive and negative samples are directly learnable end-to-end with the encoder. We show that the positive and negative samples can be cooperatively and adversarially learned by minimizing and maximizing the contrastive loss, respectively. This yields cooperative positives and adversarial negatives with respect to the encoder, which are updated to continuously track the learned representation of the query anchors over mini-batches. The proposed method achieves 71.3% and 75.3% in top-1 accuracy respectively over 200 and 800 epochs of pre-training ResNet-50 backbone on ImageNet1K without tricks such as multi-crop or stronger augmentations. With Multi-Crop, it can be further boosted into 75.7%. The source code and pre-trained model are released in https://github.com/maple-research-lab/caco.

Tingting Liao, Xiaomei Zhang, Yuliang Xiu et al. · tsinghua

This paper presents a framework for efficient 3D clothed avatar reconstruction. By combining the advantages of the high accuracy of optimization-based methods and the efficiency of learning-based methods, we propose a coarse-to-fine way to realize a high-fidelity clothed avatar reconstruction (CAR) from a single image. At the first stage, we use an implicit model to learn the general shape in the canonical space of a person in a learning-based way, and at the second stage, we refine the surface detail by estimating the non-rigid deformation in the posed space in an optimization way. A hyper-network is utilized to generate a good initialization so that the convergence o f the optimization process is greatly accelerated. Extensive experiments on various datasets show that the proposed CAR successfully produces high-fidelity avatars for arbitrarily clothed humans in real scenes.

Xiao Wang, Ying Wang, Ziwei Xuan et al.

Unsupervised learning of vision transformers seeks to pretrain an encoder via pretext tasks without labels. Among them is the Masked Image Modeling (MIM) aligned with pretraining of language transformers by predicting masked patches as a pretext task. A criterion in unsupervised pretraining is the pretext task needs to be sufficiently hard to prevent the transformer encoder from learning trivial low-level features not generalizable well to downstream tasks. For this purpose, we propose an Adversarial Positional Embedding (AdPE) approach -- It distorts the local visual structures by perturbing the position encodings so that the learned transformer cannot simply use the locally correlated patches to predict the missing ones. We hypothesize that it forces the transformer encoder to learn more discriminative features in a global context with stronger generalizability to downstream tasks. We will consider both absolute and relative positional encodings, where adversarial positions can be imposed both in the embedding mode and the coordinate mode. We will also present a new MAE+ baseline that brings the performance of the MIM pretraining to a new level with the AdPE. The experiments demonstrate that our approach can improve the fine-tuning accuracy of MAE by $0.8\%$ and $0.4\%$ over 1600 epochs of pretraining ViT-B and ViT-L on Imagenet1K. For the transfer learning task, it outperforms the MAE with the ViT-B backbone by $2.6\%$ in mIoU on ADE20K, and by $3.2\%$ in AP$^{bbox}$ and $1.6\%$ in AP$^{mask}$ on COCO, respectively. These results are obtained with the AdPE being a pure MIM approach that does not use any extra models or external datasets for pretraining. The code is available at https://github.com/maple-research-lab/AdPE.

Benzhi Wang, Yang Yang, Jinlin Wu et al.

Weakly supervised person search aims to jointly detect and match persons with only bounding box annotations. Existing approaches typically focus on improving the features by exploring relations of persons. However, scale variation problem is a more severe obstacle and under-studied that a person often owns images with different scales (resolutions). On the one hand, small-scale images contain less information of a person, thus affecting the accuracy of the generated pseudo labels. On the other hand, the similarity of cross-scale images is often smaller than that of images with the same scale for a person, which will increase the difficulty of matching. In this paper, we address this problem by proposing a novel one-step framework, named Self-similarity driven Scale-invariant Learning (SSL). Scale invariance can be explored based on the self-similarity prior that it shows the same statistical properties of an image at different scales. To this end, we introduce a Multi-scale Exemplar Branch to guide the network in concentrating on the foreground and learning scale-invariant features by hard exemplars mining. To enhance the discriminative power of the features in an unsupervised manner, we introduce a dynamic multi-label prediction which progressively seeks true labels for training. It is adaptable to different types of unlabeled data and serves as a compensation for clustering based strategy. Experiments on PRW and CUHK-SYSU databases demonstrate the effectiveness of our method.

Xueji Fang, Liyuan Ma, Jianhao Zeng et al.

Diffusion transformer (DiT) has been widely adopted in the generative diffusion field, advancing the denoising of query tokens through attention and Feed-Forward (\text{FFN}) layers. FFN actually acts as the key-value vocabulary for decoding visual contents where the value embeds the visual semantical knowledge. We present that focusing on critical query tokens corresponding to more complex details and encouraging the model to improve these tokens is essential for fine-grained visual generation. To this end, we propose FocusDiT, which applies a Masking scheme to focus on critical query tokens that are exclusively fed into FFN. The masked queries can retrieve visual tokens from the FFN vocabularies, and use them to decode their visual details. Extensive text-to-image experiments validate the effectiveness of token masking in enhancing generative performance.

Liyuan Ma, Xueji Fang, Guo-Jun Qi

Image customization learns target subjects from reference concept images and generates conditioned images per text prompts, mainly modifying styles or backgrounds. Prevailing methods adopt fine-tuning to pack diverse concept attributes into a unified latent embedding, yet entangled attributes hinder elimination of irrelevant disturbances from style and background. To address this issue, we propose Equilibrated Diffusion, a frequency-driven approach that disentangles tangled concept features for balanced customization and consistent text-visual matching. Unlike conventional methods learning full concepts with shared embeddings and unified tuning, our work utilizes the inherent link between image frequency components and semantics: low frequencies represent subject content and high frequencies correspond to styles. We decompose concepts in frequency space and optimize each embedding independently. This separate optimization enables the denoiser to capture style detached from subject identity and generalize better to unseen stylistic prompts. Merging multi-frequency embeddings preserves the model's original spatial customization ability. We further deploy mask-guided diffusion to restrict irrelevant background changes and boost text alignment. Residual Reference Attention (RRA) is inserted into spatial attention to retain subject structure and identity consistency. Experiments prove Equilibrated Diffusion exceeds mainstream baselines on subject fidelity and text adherence, verifying our method's superiority.

Na Zhang, Xudong Liu, Xin Li et al.

Semantic face image manipulation has received increasing attention in recent years. StyleGAN-based approaches to face morphing are among the leading techniques; however, they often suffer from noticeable blurring and artifacts as a result of the uniform attention in the latent feature space. In this paper, we propose to develop a transformer-based alternative to face morphing and demonstrate its superiority to StyleGAN-based methods. Our contributions are threefold. First, inspired by GANformer, we introduce a bipartite structure to exploit long-range interactions in face images for iterative propagation of information from latent variables to salient facial features. Special loss functions are designed to support the optimization of face morphing. Second, we extend the study of transformer-based face morphing to demorphing by presenting an effective defense strategy with access to a reference image using the same generator of MorphGANFormer. Such demorphing is conceptually similar to unmixing of hyperspectral images but operates in the latent (instead of pixel) space. Third, for the first time, we address a fundamental issue of vulnerability-detectability trade-off for face morphing studies. It is argued that neither doppelganger norrandom pair selection is optimal, and a Lagrangian multiplier-based approach should be used to achieve an improved trade-off between recognition vulnerability and attack detectability.

Sanyi Zhang, Xiaochun Cao, Guo-Jun Qi et al.

Most state-of-the-art instance-level human parsing models adopt two-stage anchor-based detectors and, therefore, cannot avoid the heuristic anchor box design and the lack of analysis on a pixel level. To address these two issues, we have designed an instance-level human parsing network which is anchor-free and solvable on a pixel level. It consists of two simple sub-networks: an anchor-free detection head for bounding box predictions and an edge-guided parsing head for human segmentation. The anchor-free detector head inherits the pixel-like merits and effectively avoids the sensitivity of hyper-parameters as proved in object detection applications. By introducing the part-aware boundary clue, the edge-guided parsing head is capable to distinguish adjacent human parts from among each other up to 58 parts in a single human instance, even overlapping instances. Meanwhile, a refinement head integrating box-level score and part-level parsing quality is exploited to improve the quality of the parsing results. Experiments on two multiple human parsing datasets (i.e., CIHP and LV-MHP-v2.0) and one video instance-level human parsing dataset (i.e., VIP) show that our method achieves the best global-level and instance-level performance over state-of-the-art one-stage top-down alternatives.

Xianpeng Liu, Ce Zheng, Kelvin Cheng et al.

The main challenge of monocular 3D object detection is the accurate localization of 3D center. Motivated by a new and strong observation that this challenge can be remedied by a 3D-space local-grid search scheme in an ideal case, we propose a stage-wise approach, which combines the information flow from 2D-to-3D (3D bounding box proposal generation with a single 2D image) and 3D-to-2D (proposal verification by denoising with 3D-to-2D contexts) in a top-down manner. Specifically, we first obtain initial proposals from off-the-shelf backbone monocular 3D detectors. Then, we generate a 3D anchor space by local-grid sampling from the initial proposals. Finally, we perform 3D bounding box denoising at the 3D-to-2D proposal verification stage. To effectively learn discriminative features for denoising highly overlapped proposals, this paper presents a method of using the Perceiver I/O model to fuse the 3D-to-2D geometric information and the 2D appearance information. With the encoded latent representation of a proposal, the verification head is implemented with a self-attention module. Our method, named as MonoXiver, is generic and can be easily adapted to any backbone monocular 3D detectors. Experimental results on the well-established KITTI dataset and the challenging large-scale Waymo dataset show that MonoXiver consistently achieves improvement with limited computation overhead.

Ce Zheng, Xianpeng Liu, Guo-Jun Qi et al.

Transformer architectures have achieved SOTA performance on the human mesh recovery (HMR) from monocular images. However, the performance gain has come at the cost of substantial memory and computational overhead. A lightweight and efficient model to reconstruct accurate human mesh is needed for real-world applications. In this paper, we propose a pure transformer architecture named POoling aTtention TransformER (POTTER) for the HMR task from single images. Observing that the conventional attention module is memory and computationally expensive, we propose an efficient pooling attention module, which significantly reduces the memory and computational cost without sacrificing performance. Furthermore, we design a new transformer architecture by integrating a High-Resolution (HR) stream for the HMR task. The high-resolution local and global features from the HR stream can be utilized for recovering more accurate human mesh. Our POTTER outperforms the SOTA method METRO by only requiring 7% of total parameters and 14% of the Multiply-Accumulate Operations on the Human3.6M (PA-MPJPE metric) and 3DPW (all three metrics) datasets. The project webpage is https://zczcwh.github.io/potter_page.

Guo-Jun Qi, Mubarak Shah

In this paper, we review adversarial pretraining of self-supervised deep networks including both convolutional neural networks and vision transformers. Unlike the adversarial training with access to labeled examples, adversarial pretraining is complicated as it only has access to unlabeled examples. To incorporate adversaries into pretraining models on either input or feature level, we find that existing approaches are largely categorized into two groups: memory-free instance-wise attacks imposing worst-case perturbations on individual examples, and memory-based adversaries shared across examples over iterations. In particular, we review several representative adversarial pretraining models based on Contrastive Learning (CL) and Masked Image Modeling (MIM), respectively, two popular self-supervised pretraining methods in literature. We also review miscellaneous issues about computing overheads, input-/feature-level adversaries, as well as other adversarial pretraining approaches beyond the above two groups. Finally, we discuss emerging trends and future directions about the relations between adversarial and cooperative pretraining, unifying adversarial CL and MIM pretraining, and the trade-off between accuracy and robustness in adversarial pretraining.

Ce Zheng, Matias Mendieta, Taojiannan Yang et al.

Recently, vision transformers have shown great success in a set of human reconstruction tasks such as 2D human pose estimation (2D HPE), 3D human pose estimation (3D HPE), and human mesh reconstruction (HMR) tasks. In these tasks, feature map representations of the human structural information are often extracted first from the image by a CNN (such as HRNet), and then further processed by transformer to predict the heatmaps (encodes each joint's location into a feature map with a Gaussian distribution) for HPE or HMR. However, existing transformer architectures are not able to process these feature map inputs directly, forcing an unnatural flattening of the location-sensitive human structural information. Furthermore, much of the performance benefit in recent HPE and HMR methods has come at the cost of ever-increasing computation and memory needs. Therefore, to simultaneously address these problems, we propose FeatER, a novel transformer design that preserves the inherent structure of feature map representations when modeling attention while reducing memory and computational costs. Taking advantage of FeatER, we build an efficient network for a set of human reconstruction tasks including 2D HPE, 3D HPE, and HMR. A feature map reconstruction module is applied to improve the performance of the estimated human pose and mesh. Extensive experiments demonstrate the effectiveness of FeatER on various human pose and mesh datasets. For instance, FeatER outperforms the SOTA method MeshGraphormer by requiring 5% of Params and 16% of MACs on Human3.6M and 3DPW datasets. The project webpage is https://zczcwh.github.io/feater_page/.

Zilyu Ye, Jinxiu Liu, Ruotian Peng et al.

Recent image generation models excel at creating high-quality images from brief captions. However, they fail to maintain consistency of multiple instances across images when encountering lengthy contexts. This inconsistency is largely due to in existing training datasets the absence of granular instance feature labeling in existing training datasets. To tackle these issues, we introduce Openstory++, a large-scale dataset combining additional instance-level annotations with both images and text. Furthermore, we develop a training methodology that emphasizes entity-centric image-text generation, ensuring that the models learn to effectively interweave visual and textual information. Specifically, Openstory++ streamlines the process of keyframe extraction from open-domain videos, employing vision-language models to generate captions that are then polished by a large language model for narrative continuity. It surpasses previous datasets by offering a more expansive open-domain resource, which incorporates automated captioning, high-resolution imagery tailored for instance count, and extensive frame sequences for temporal consistency. Additionally, we present Cohere-Bench, a pioneering benchmark framework for evaluating the image generation tasks when long multimodal context is provided, including the ability to keep the background, style, instances in the given context coherent. Compared to existing benchmarks, our work fills critical gaps in multi-modal generation, propelling the development of models that can adeptly generate and interpret complex narratives in open-domain environments. Experiments conducted within Cohere-Bench confirm the superiority of Openstory++ in nurturing high-quality visual storytelling models, enhancing their ability to address open-domain generation tasks. More details can be found at https://openstorypp.github.io/

Zhangsihao Yang, Mingyuan Zhou, Mengyi Shan et al.

Our paper aims to generate diverse and realistic animal motion sequences from textual descriptions, without a large-scale animal text-motion dataset. While the task of text-driven human motion synthesis is already extensively studied and benchmarked, it remains challenging to transfer this success to other skeleton structures with limited data. In this work, we design a model architecture that imitates Generative Pretraining Transformer (GPT), utilizing prior knowledge learned from human data to the animal domain. We jointly train motion autoencoders for both animal and human motions and at the same time optimize through the similarity scores among human motion encoding, animal motion encoding, and text CLIP embedding. Presenting the first solution to this problem, we are able to generate animal motions with high diversity and fidelity, quantitatively and qualitatively outperforming the results of training human motion generation baselines on animal data. Additionally, we introduce AnimalML3D, the first text-animal motion dataset with 1240 animation sequences spanning 36 different animal identities. We hope this dataset would mediate the data scarcity problem in text-driven animal motion generation, providing a new playground for the research community.

Tiancheng Li, Weijian Luo, Zhiyang Chen et al.

Proper guidance strategies are essential to achieve high-quality generation results without retraining diffusion and flow-based text-to-image models. Existing guidance either requires specific training or strong inductive biases of diffusion model networks, which potentially limits their ability and application scope. Motivated by the observation that artifact outliers can be detected by a significant decline in the density from a noisier to a cleaner noise level, we propose Self-Guidance (SG), which can significantly improve the quality of the generated image by suppressing the generation of low-quality samples. The biggest difference from existing guidance is that SG only relies on the sampling score function of the original diffusion or flow model at different noise levels, with no need for any tricky and expensive guidance-specific training. This makes SG highly flexible to be used in a plug-and-play manner by any diffusion or flow models. We also introduce an efficient variant of SG, named SG-prev, which reuses the output from the immediately previous diffusion step to avoid additional forward passes of the diffusion network.We conduct extensive experiments on text-to-image and text-to-video generation with different architectures, including UNet and transformer models. With open-sourced diffusion models such as Stable Diffusion 3.5 and FLUX, SG exceeds existing algorithms on multiple metrics, including both FID and Human Preference Score. SG-prev also achieves strong results over both the baseline and the SG, with 50 percent more efficiency. Moreover, we find that SG and SG-prev both have a surprisingly positive effect on the generation of physiologically correct human body structures such as hands, faces, and arms, showing their ability to eliminate human body artifacts with minimal efforts. We have released our code at https://github.com/maple-research-lab/Self-Guidance.

Xueji Fang, Liyuan Ma, Zhiyang Chen et al.

Recent advances in text-to-video generation, particularly with autoregressive models, have enabled the synthesis of high-quality videos depicting individual scenes. However, extending these models to generate long, cross-scene videos remains a significant challenge. As the context length grows during autoregressive decoding, computational costs rise sharply, and the model's ability to maintain consistency and adhere to evolving textual prompts deteriorates. We introduce InfLVG, an inference-time framework that enables coherent long video generation without requiring additional long-form video data. InfLVG leverages a learnable context selection policy, optimized via Group Relative Policy Optimization (GRPO), to dynamically identify and retain the most semantically relevant context throughout the generation process. Instead of accumulating the entire generation history, the policy ranks and selects the top-$K$ most contextually relevant tokens, allowing the model to maintain a fixed computational budget while preserving content consistency and prompt alignment. To optimize the policy, we design a hybrid reward function that jointly captures semantic alignment, cross-scene consistency, and artifact reduction. To benchmark performance, we introduce the Cross-scene Video Benchmark (CsVBench) along with an Event Prompt Set (EPS) that simulates complex multi-scene transitions involving shared subjects and varied actions/backgrounds. Experimental results show that InfLVG can extend video length by up to 9$\times$, achieving strong consistency and semantic fidelity across scenes. Our code is available at https://github.com/MAPLE-AIGC/InfLVG.

Guoan Xu, Yang Xiao, Guangwei Gao et al.

Multimodal semantic segmentation has emerged as a powerful paradigm for enhancing scene understanding by leveraging complementary information from multiple sensing modalities (e.g., RGB, depth, and thermal). However, existing cross-modal fusion methods often implicitly assume that all modalities are equally reliable, which can lead to feature degradation when auxiliary modalities are noisy, misaligned, or incomplete. In this paper, we revisit cross-modal fusion from the perspective of modality reliability and propose a novel framework termed the Reliability-aware Self-Gated State Space Model (RSGMamba). At the core of our method is the Reliability-aware Self-Gated Mamba Block (RSGMB), which explicitly models modality reliability and dynamically regulates cross-modal interactions through a self-gating mechanism. Unlike conventional fusion strategies that indiscriminately exchange information across modalities, RSGMB enables reliability-aware feature selection and enhancing informative feature aggregation. In addition, a lightweight Local Cross-Gated Modulation (LCGM) is incorporated to refine fine-grained spatial details, complementing the global modeling capability of RSGMB. Extensive experiments demonstrate that RSGMamba achieves state-of-the-art performance on both RGB-D and RGB-T semantic segmentation benchmarks, resulting 58.8% / 54.0% mIoU on NYUDepth V2 and SUN-RGBD (+0.4% / +0.7% over prior best), and 61.1% / 88.9% mIoU on MFNet and PST900 (up to +1.6%), with only 48.6M parameters, thereby validating the effectiveness and superiority of the proposed approach.

Zemin Huang, Yuhang Wang, Zhiyang Chen et al.

Mask-based Diffusion Language Models (DLMs) struggle to revise incorrect tokens: once a token is generated, it typically remains fixed. The key challenge is to identify potential errors in the inputs. In this paper, we propose \emph{\underline{Rem}asking-\underline{e}nabled \underline{Di}ffusion Language Model (RemeDi}, a mask-based DLM that introduces \emph{remasking} as another fundamental mechanism, enabling more flexible text refinement in diffusion-based text generation. To achieve this, RemeDi jointly predicts token distributions and per-token confidence scores at each step. The confidence scores determine which tokens to be unmasked after the current step, allowing the model to identify tokens with low quality and remask them. These remasked tokens can be resampled with richer context in subsequent steps. We design a remask-aware pipeline to train this ability, including supervised fine-tuning which teaches the model to detect and remask incorrect tokens in addition to predict mask tokens, and reinforcement learning which optimizes full generation trajectories toward higher rewards. Experiments show that RemeDi achieves the state-of-the-art results among open-source DLMs on multiple datasets.

Xiao Wang, Guo-Jun Qi

Representation learning has significantly been developed with the advance of contrastive learning methods. Most of those methods have benefited from various data augmentations that are carefully designated to maintain their identities so that the images transformed from the same instance can still be retrieved. However, those carefully designed transformations limited us to further explore the novel patterns exposed by other transformations. Meanwhile, as found in our experiments, the strong augmentations distorted the images' structures, resulting in difficult retrieval. Thus, we propose a general framework called Contrastive Learning with Stronger Augmentations~(CLSA) to complement current contrastive learning approaches. Here, the distribution divergence between the weakly and strongly augmented images over the representation bank is adopted to supervise the retrieval of strongly augmented queries from a pool of instances. Experiments on the ImageNet dataset and downstream datasets showed the information from the strongly augmented images can significantly boost the performance. For example, CLSA achieves top-1 accuracy of 76.2% on ImageNet with a standard ResNet-50 architecture with a single-layer classifier fine-tuned, which is almost the same level as 76.5% of supervised results. The code and pre-trained models are available in https://github.com/maple-research-lab/CLSA.

Si Chen, Mostafa Kahla, Ruoxi Jia et al.

Model inversion (MI) attacks are aimed at reconstructing training data from model parameters. Such attacks have triggered increasing concerns about privacy, especially given a growing number of online model repositories. However, existing MI attacks against deep neural networks (DNNs) have large room for performance improvement. We present a novel inversion-specific GAN that can better distill knowledge useful for performing attacks on private models from public data. In particular, we train the discriminator to differentiate not only the real and fake samples but the soft-labels provided by the target model. Moreover, unlike previous work that directly searches for a single data point to represent a target class, we propose to model a private data distribution for each target class. Our experiments show that the combination of these techniques can significantly boost the success rate of the state-of-the-art MI attacks by 150%, and generalize better to a variety of datasets and models. Our code is available at https://github.com/SCccc21/Knowledge-Enriched-DMI.

Xiao Wang, Daisuke Kihara, Jiebo Luo et al.

Deep neural networks have been successfully applied to many real-world applications. However, such successes rely heavily on large amounts of labeled data that is expensive to obtain. Recently, many methods for semi-supervised learning have been proposed and achieved excellent performance. In this study, we propose a new EnAET framework to further improve existing semi-supervised methods with self-supervised information. To our best knowledge, all current semi-supervised methods improve performance with prediction consistency and confidence ideas. We are the first to explore the role of {\bf self-supervised} representations in {\bf semi-supervised} learning under a rich family of transformations. Consequently, our framework can integrate the self-supervised information as a regularization term to further improve {\it all} current semi-supervised methods. In the experiments, we use MixMatch, which is the current state-of-the-art method on semi-supervised learning, as a baseline to test the proposed EnAET framework. Across different datasets, we adopt the same hyper-parameters, which greatly improves the generalization ability of the EnAET framework. Experiment results on different datasets demonstrate that the proposed EnAET framework greatly improves the performance of current semi-supervised algorithms. Moreover, this framework can also improve {\bf supervised learning} by a large margin, including the extremely challenging scenarios with only 10 images per class. The code and experiment records are available in \url{https://github.com/maple-research-lab/EnAET}.

Xiang Gao, Wei Hu, Guo-Jun Qi

Recent advances in Graph Convolutional Neural Networks (GCNNs) have shown their efficiency for non-Euclidean data on graphs, which often require a large amount of labeled data with high cost. It it thus critical to learn graph feature representations in an unsupervised manner in practice. To this end, we propose a novel unsupervised learning of Graph Transformation Equivariant Representations (GraphTER), aiming to capture intrinsic patterns of graph structure under both global and local transformations. Specifically, we allow to sample different groups of nodes from a graph and then transform them node-wise isotropically or anisotropically. Then, we self-train a representation encoder to capture the graph structures by reconstructing these node-wise transformations from the feature representations of the original and transformed graphs. In experiments, we apply the learned GraphTER to graphs of 3D point cloud data, and results on point cloud segmentation/classification show that GraphTER significantly outperforms state-of-the-art unsupervised approaches and pushes greatly closer towards the upper bound set by the fully supervised counterparts. The code is available at: https://github.com/gyshgx868/graph-ter.

Yuhui Xu, Lingxi Xie, Xiaopeng Zhang et al.

Differentiable architecture search (DARTS) provided a fast solution in finding effective network architectures, but suffered from large memory and computing overheads in jointly training a super-network and searching for an optimal architecture. In this paper, we present a novel approach, namely, Partially-Connected DARTS, by sampling a small part of super-network to reduce the redundancy in exploring the network space, thereby performing a more efficient search without comprising the performance. In particular, we perform operation search in a subset of channels while bypassing the held out part in a shortcut. This strategy may suffer from an undesired inconsistency on selecting the edges of super-net caused by sampling different channels. We alleviate it using edge normalization, which adds a new set of edge-level parameters to reduce uncertainty in search. Thanks to the reduced memory cost, PC-DARTS can be trained with a larger batch size and, consequently, enjoys both faster speed and higher training stability. Experimental results demonstrate the effectiveness of the proposed method. Specifically, we achieve an error rate of 2.57% on CIFAR10 with merely 0.1 GPU-days for architecture search, and a state-of-the-art top-1 error rate of 24.2% on ImageNet (under the mobile setting) using 3.8 GPU-days for search. Our code has been made available at: https://github.com/yuhuixu1993/PC-DARTS.

Weijian Luo, Zemin Huang, Zhengyang Geng et al.

Despite their strong performances on many generative tasks, diffusion models require a large number of sampling steps in order to generate realistic samples. This has motivated the community to develop effective methods to distill pre-trained diffusion models into more efficient models, but these methods still typically require few-step inference or perform substantially worse than the underlying model. In this paper, we present Score Implicit Matching (SIM) a new approach to distilling pre-trained diffusion models into single-step generator models, while maintaining almost the same sample generation ability as the original model as well as being data-free with no need of training samples for distillation. The method rests upon the fact that, although the traditional score-based loss is intractable to minimize for generator models, under certain conditions we can efficiently compute the gradients for a wide class of score-based divergences between a diffusion model and a generator. SIM shows strong empirical performances for one-step generators: on the CIFAR10 dataset, it achieves an FID of 2.06 for unconditional generation and 1.96 for class-conditional generation. Moreover, by applying SIM to a leading transformer-based diffusion model, we distill a single-step generator for text-to-image (T2I) generation that attains an aesthetic score of 6.42 with no performance decline over the original multi-step counterpart, clearly outperforming the other one-step generators including SDXL-TURBO of 5.33, SDXL-LIGHTNING of 5.34 and HYPER-SDXL of 5.85. We will release this industry-ready one-step transformer-based T2I generator along with this paper.

Zemin Huang, Zhiyang Chen, Zijun Wang et al.

We introduce the Diffusion Chain of Lateral Thought (DCoLT), a reasoning framework for diffusion language models. DCoLT treats each intermediate step in the reverse diffusion process as a latent "thinking" action and optimizes the entire reasoning trajectory to maximize the reward on the correctness of the final answer with outcome-based Reinforcement Learning (RL). Unlike traditional Chain-of-Thought (CoT) methods that follow a causal, linear thinking process, DCoLT allows bidirectional, non-linear reasoning with no strict rule on grammatical correctness amid its intermediate steps of thought. We implement DCoLT on two representative Diffusion Language Models (DLMs). First, we choose SEDD as a representative continuous-time discrete diffusion model, where its concrete score derives a probabilistic policy to maximize the RL reward over the entire sequence of intermediate diffusion steps. We further consider the discrete-time masked diffusion language model -- LLaDA, and find that the order to predict and unmask tokens plays an essential role to optimize its RL action resulting from the ranking-based Unmasking Policy Module (UPM) defined by the Plackett-Luce model. Experiments on both math and code generation tasks show that using only public data and 16 H800 GPUs, DCoLT-reinforced DLMs outperform other DLMs trained by SFT or RL or even both. Notably, DCoLT-reinforced LLaDA boosts its reasoning accuracy by +9.8%, +5.7%, +11.4%, +19.5% on GSM8K, MATH, MBPP, and HumanEval.

Zemin Huang, Zhengyang Geng, Weijian Luo et al.

In the realm of Artificial Intelligence Generated Content (AIGC), flow-matching models have emerged as a powerhouse, achieving success due to their robust theoretical underpinnings and solid ability for large-scale generative modeling. These models have demonstrated state-of-the-art performance, but their brilliance comes at a cost. The process of sampling from these models is notoriously demanding on computational resources, as it necessitates the use of multi-step numerical ordinary differential equations (ODEs). Against this backdrop, this paper presents a novel solution with theoretical guarantees in the form of Flow Generator Matching (FGM), an innovative approach designed to accelerate the sampling of flow-matching models into a one-step generation, while maintaining the original performance. On the CIFAR10 unconditional generation benchmark, our one-step FGM model achieves a new record Fréchet Inception Distance (FID) score of 3.08 among few-step flow-matching-based models, outperforming original 50-step flow-matching models. Furthermore, we use the FGM to distill the Stable Diffusion 3, a leading text-to-image flow-matching model based on the MM-DiT architecture. The resulting MM-DiT-FGM one-step text-to-image model demonstrates outstanding industry-level performance. When evaluated on the GenEval benchmark, MM-DiT-FGM has delivered remarkable generating qualities, rivaling other multi-step models in light of the efficiency of a single generation step.

Zilyu Ye, Zhiyang Chen, Tiancheng Li et al.

Diffusion and flow matching models have achieved remarkable success in text-to-image generation. However, these models typically rely on the predetermined denoising schedules for all prompts. The multi-step reverse diffusion process can be regarded as a kind of chain-of-thought for generating high-quality images step by step. Therefore, diffusion models should reason for each instance to adaptively determine the optimal noise schedule, achieving high generation quality with sampling efficiency. In this paper, we introduce the Time Prediction Diffusion Model (TPDM) for this. TPDM employs a plug-and-play Time Prediction Module (TPM) that predicts the next noise level based on current latent features at each denoising step. We train the TPM using reinforcement learning to maximize a reward that encourages high final image quality while penalizing excessive denoising steps. With such an adaptive scheduler, TPDM not only generates high-quality images that are aligned closely with human preferences but also adjusts diffusion time and the number of denoising steps on the fly, enhancing both performance and efficiency. With Stable Diffusion 3 Medium architecture, TPDM achieves an aesthetic score of 5.44 and a human preference score (HPS) of 29.59, while using around 50% fewer denoising steps to achieve better performance.

Bingwen Zhu, Fanyi Wang, Tianyi Lu et al.

Image-to-video (I2V) generation aims to create a video sequence from a single image, which requires high temporal coherence and visual fidelity. However, existing approaches suffer from inconsistency of character appearances and poor preservation of fine details. Moreover, they require a large amount of video data for training, which can be computationally demanding. To address these limitations, we propose PoseAnimate, a novel zero-shot I2V framework for character animation. PoseAnimate contains three key components: 1) a Pose-Aware Control Module (PACM) that incorporates diverse pose signals into text embeddings, to preserve character-independent content and maintain precise alignment of actions. 2) a Dual Consistency Attention Module (DCAM) that enhances temporal consistency and retains character identity and intricate background details. 3) a Mask-Guided Decoupling Module (MGDM) that refines distinct feature perception abilities, improving animation fidelity by decoupling the character and background. We also propose a Pose Alignment Transition Algorithm (PATA) to ensure smooth action transition. Extensive experiment results demonstrate that our approach outperforms the state-of-the-art training-based methods in terms of character consistency and detail fidelity. Moreover, it maintains a high level of temporal coherence throughout the generated animations.

Yuming Qiao, Fanyi Wang, Jingwen Su et al.

Image editing approaches with diffusion models have been rapidly developed, yet their applicability are subject to requirements such as specific editing types (e.g., foreground or background object editing, style transfer), multiple conditions (e.g., mask, sketch, caption), and time consuming fine-tuning of diffusion models. For alleviating these limitations and realizing efficient real image editing, we propose a novel editing technique that only requires an input image and target text for various editing types including non-rigid edits without fine-tuning diffusion model. Our method contains three novelties:(I) Target-text Inversion Schedule (TTIS) is designed to fine-tune the input target text embedding to achieve fast image reconstruction without image caption and acceleration of convergence.(II) Progressive Transition Scheme applies progressive linear interpolation between target text embedding and its fine-tuned version to generate transition embedding for maintaining non-rigid editing capability.(III) Balanced Attention Module (BAM) balances the tradeoff between textual description and image semantics.By the means of combining self-attention map from reconstruction process and cross-attention map from transition process, the guidance of target text embeddings in diffusion process is optimized.In order to demonstrate editing capability, effectiveness and efficiency of the proposed BARET, we have conducted extensive qualitative and quantitative experiments. Moreover, results derived from user study and ablation study further prove the superiority over other methods.

Tiancheng Li, Yuhang Wang, Zhiyang Chen et al.

Prompt evolution algorithms offer a powerful paradigm for enhancing AI systems based on closed-source models, while few work explores whether aggregating results from multiple prompts to reach a consensus can further advance the system capability boundary. In this paper, we introduce Consensus-Evolve (C-Evolve), an evolutionary algorithm that discovers a group of prompts whose aggregated outputs after majority voting achieve optimal performance. More specifically, C-Evolve employs an island-based evolutionary algorithm to maintain population diversity, and prompts from distinct islands are selected to form groups to aggregate their outputs. The key difference from single individual evolution is a voting score, which evaluates each individual prompt's contribution within groups. We take this as the fitness score for evolution instead of individual performance. Consequently, C-Evolve is more likely to produce and maintain prompts with higher potential to form a high-performing group and eliminate low-performing ones, gradually improving the group performance after reaching consensus. Our method achieves state-of-the-art performance across a wide range of tasks, including both open-ended tasks like HotpotQA and closed-ended tasks like MATH. On Qwen3-8B, C-Evolve achieves 70.67% on HotpotQA and 43.88% on IFBench, which are 4.95% and 2.73% higher than GEPA, respectively. For GPT-4.1-mini, the accuracy on IFBench is further improved to 47.96% and reaches 95.33% in the MATH benchmark. These results demonstrate the C-Evolve's competitive performance.

Xuehao Gao, Yang Yang, Shaoyi Du et al.

This paper explores a cross-modality synthesis task that infers 3D human-object interactions (HOIs) from a given text-based instruction. Existing text-to-HOI synthesis methods mainly deploy a direct mapping from texts to object-specific 3D body motions, which may encounter a performance bottleneck since the huge cross-modality gap. In this paper, we observe that those HOI samples with the same interaction intention toward different targets, e.g., "lift a chair" and "lift a cup", always encapsulate similar action-specific body motion patterns while characterizing different object-specific interaction styles. Thus, learning effective action-specific motion priors and object-specific interaction priors is crucial for a text-to-HOI model and dominates its performances on text-HOI semantic consistency and body-object interaction realism. In light of this, we propose a novel body pose generation strategy for the text-to-HOI task: infer object-agnostic canonical body action first and then enrich object-specific interaction styles. Specifically, the first canonical body action inference stage focuses on learning intra-class shareable body motion priors and mapping given text-based semantics to action-specific canonical 3D body motions. Then, in the object-specific interaction inference stage, we focus on object affordance learning and enrich object-specific interaction styles on an inferred action-specific body motion basis. Extensive experiments verify that our proposed text-to-HOI synthesis system significantly outperforms other SOTA methods on three large-scale datasets with better semantic consistency and interaction realism performances.

Xuehao Gao, Yang Yang, Shaoyi Du et al.

As two intimate reciprocal tasks, scene-aware human motion synthesis and analysis require a joint understanding between multiple modalities, including 3D body motions, 3D scenes, and textual descriptions. In this paper, we integrate these two paired processes into a Co-Evolving Synthesis-Analysis (CESA) pipeline and mutually benefit their learning. Specifically, scene-aware text-to-human synthesis generates diverse indoor motion samples from the same textual description to enrich human-scene interaction intra-class diversity, thus significantly benefiting training a robust human motion analysis system. Reciprocally, human motion analysis would enforce semantic scrutiny on each synthesized motion sample to ensure its semantic consistency with the given textual description, thus improving realistic motion synthesis. Considering that real-world indoor human motions are goal-oriented and path-guided, we propose a cascaded generation strategy that factorizes text-driven scene-specific human motion generation into three stages: goal inferring, path planning, and pose synthesizing. Coupling CESA with this powerful cascaded motion synthesis model, we jointly improve realistic human motion synthesis and robust human motion analysis in 3D scenes.

Guoan Xu, Jiaming Chen, Wenfeng Huang et al.

The Vision Transformer (ViT) has achieved notable success in computer vision, with its variants extensively validated across various downstream tasks, including semantic segmentation. However, designed as general-purpose visual encoders, ViT backbones often overlook the specific needs of task decoders, revealing opportunities to design decoders tailored to efficient semantic segmentation. This paper proposes Strip Cross-Attention (SCASeg), an innovative decoder head explicitly designed for semantic segmentation. Instead of relying on the simple conventional skip connections, we employ lateral connections between the encoder and decoder stages, using encoder features as Queries for the cross-attention modules. Additionally, we introduce a Cross-Layer Block that blends hierarchical feature maps from different encoder and decoder stages to create a unified representation for Keys and Values. To further boost computational efficiency, SCASeg compresses queries and keys into strip-like patterns to optimize memory usage and inference speed over the traditional vanilla cross-attention. Moreover, the Cross-Layer Block incorporates the local perceptual strengths of convolution, enabling SCASeg to capture both global and local context dependencies across multiple layers. This approach facilitates effective feature interaction at different scales, improving the overall performance. Experiments show that the adaptable decoder of SCASeg produces competitive performance across different setups, surpassing leading segmentation architectures on all benchmark datasets, including ADE20K, Cityscapes, COCO-Stuff 164k, and Pascal VOC2012, even under varying computational limitations.

Zesheng Yang, Xi Jiang, Bingzhang Hu et al.

Current vision-language detection and grounding models predominantly focus on prompts with positive semantics and often struggle to accurately interpret and ground complex expressions containing negative semantics. A key reason for this limitation is the lack of high-quality training data that explicitly captures discriminative negative samples and negation-aware language descriptions. To address this challenge, we introduce D-Negation, a new dataset that provides objects annotated with both positive and negative semantic descriptions. Building upon the observation that negation reasoning frequently appears in natural language, we further propose a grouped opposition-based learning framework that learns negation-aware representations from limited samples. Specifically, our method organizes opposing semantic descriptions from D-Negation into structured groups and formulates two complementary loss functions that encourage the model to reason about negation and semantic qualifiers. We integrate the proposed dataset and learning strategy into a state-of-the-art language-based grounding model. By fine-tuning fewer than 10 percent of the model parameters, our approach achieves improvements of up to 4.4 mAP and 5.7 mAP on positive and negative semantic evaluations, respectively. These results demonstrate that explicitly modeling negation semantics can substantially enhance the robustness and localization accuracy of vision-language grounding models.

Guoan Xu, Yang Xiao, Wenjing Jia et al.

While recent semantic segmentation networks heavily rely on powerful pretrained encoders, most employ simplistic decoders, leading to suboptimal trade-offs between semantic context and fine-grained detail preservation. To address this, we propose a novel decoder architecture, WaveSeg, which jointly optimizes feature refinement in spatial and wavelet domains. Specifically, high-frequency components are first learned from input images as explicit priors to reinforce boundary details at early stages. A multi-scale fusion mechanism, Dual Domain Operation (DDO), is then applied, and the novel Spectrum Decomposition Attention (SDA) block is proposed, which is developed to leverage Mamba's linear-complexity long-range modeling to enhance high-frequency structural details. Meanwhile, reparameterized convolutions are applied to preserve low-frequency semantic integrity in the wavelet domain. Finally, a residual-guided fusion integrates multi-scale features with boundary-aware representations at native resolution, producing semantically and structurally rich feature maps. Extensive experiments on standard benchmarks demonstrate that WaveSeg, leveraging wavelet-domain frequency prior with Mamba-based attention, consistently outperforms state-of-the-art approaches both quantitatively and qualitatively, achieving efficient and precise segmentation.

Han Zhou, Jinjin Cao, Liyuan Ma et al.

Learning diverse manipulation skills for real-world robots is severely bottlenecked by the reliance on costly and hard-to-scale teleoperated demonstrations. While human videos offer a scalable alternative, effectively transferring manipulation knowledge is fundamentally hindered by the significant morphological gap between human and robotic embodiments. To address this challenge and facilitate skill transfer from human to robot, we introduce Traj2Action,a novel framework that bridges this embodiment gap by using the 3D trajectory of the operational endpoint as a unified intermediate representation, and then transfers the manipulation knowledge embedded in this trajectory to the robot's actions. Our policy first learns to generate a coarse trajectory, which forms an high-level motion plan by leveraging both human and robot data. This plan then conditions the synthesis of precise, robot-specific actions (e.g., orientation and gripper state) within a co-denoising framework. Extensive real-world experiments on a Franka robot demonstrate that Traj2Action boosts the performance by up to 27% and 22.25% over $π_0$ baseline on short- and long-horizon real-world tasks, and achieves significant gains as human data scales in robot policy learning. Our project website, featuring code and video demonstrations, is available at https://anonymous.4open.science/w/Traj2Action-4A45/.

Siyu Xu, Wenjie Li, Guangwei Gao et al.

Face super-resolution (FSR) under limited computational costs remains an open problem. Existing approaches typically treat all facial pixels equally, resulting in suboptimal allocation of computational resources and degraded FSR performance. CNN is relatively sensitive to high-frequency facial features, such as component contours and facial outlines. Meanwhile, Mamba excels at capturing low-frequency features like facial color and fine-grained texture, and does so with lower complexity than Transformers. Motivated by these observations, we propose FADPNet, a Frequency-Aware Dual-Path Network that decomposes facial features into low- and high-frequency components and processes them via dedicated branches. For low-frequency regions, we introduce a Mamba-based Low-Frequency Enhancement Block (LFEB), which combines state-space attention with squeeze-and-excitation operations to extract low-frequency global interactions and emphasize informative channels. For high-frequency regions, we design a CNN-based Deep Position-Aware Attention (DPA) module to enhance spatially-dependent structural details, complemented by a lightweight High-Frequency Refinement (HFR) module that further refines frequency-specific representations. Through the above designs, our method achieves an excellent balance between FSR quality and model efficiency, outperforming existing approaches.

Guoan Xu, Wenfeng Huang, Wenjing Jia et al.

Vision Transformer (ViT) has made significant advancements in computer vision, thanks to its token mixer's sophisticated ability to capture global dependencies between all tokens. However, the quadratic growth in computational demands as the number of tokens increases limits its practical efficiency. Although recent methods have combined the strengths of convolutions and self-attention to achieve better trade-offs, the expensive pairwise token affinity and complex matrix operations inherent in self-attention remain a bottleneck. To address this challenge, we propose S2AFormer, an efficient Vision Transformer architecture featuring novel Strip Self-Attention (SSA). We design simple yet effective Hybrid Perception Blocks (HPBs) to effectively integrate the local perception capabilities of CNNs with the global context modeling of Transformer's attention mechanisms. A key innovation of SSA lies in its reducing the spatial dimensions of $K$ and $V$ while compressing the channel dimensions of $Q$ and $K$. This design significantly reduces computational overhead while preserving accuracy, striking an optimal balance between efficiency and effectiveness. We evaluate the robustness and efficiency of S2AFormer through extensive experiments on multiple vision benchmarks, including ImageNet-1k for image classification, ADE20k for semantic segmentation, and COCO for object detection and instance segmentation. Results demonstrate that S2AFormer achieves significant accuracy gains with superior efficiency in both GPU and non-GPU environments, making it a strong candidate for efficient vision Transformers.

Sanyi Zhang, Xiaochun Cao, Rui Wang et al.

Human parsing aims to segment each pixel of the human image with fine-grained semantic categories. However, current human parsers trained with clean data are easily confused by numerous image corruptions such as blur and noise. To improve the robustness of human parsers, in this paper, we construct three corruption robustness benchmarks, termed LIP-C, ATR-C, and Pascal-Person-Part-C, to assist us in evaluating the risk tolerance of human parsing models. Inspired by the data augmentation strategy, we propose a novel heterogeneous augmentation-enhanced mechanism to bolster robustness under commonly corrupted conditions. Specifically, two types of data augmentations from different views, i.e., image-aware augmentation and model-aware image-to-image transformation, are integrated in a sequential manner for adapting to unforeseen image corruptions. The image-aware augmentation can enrich the high diversity of training images with the help of common image operations. The model-aware augmentation strategy that improves the diversity of input data by considering the model's randomness. The proposed method is model-agnostic, and it can plug and play into arbitrary state-of-the-art human parsing frameworks. The experimental results show that the proposed method demonstrates good universality which can improve the robustness of the human parsing models and even the semantic segmentation models when facing various image common corruptions. Meanwhile, it can still obtain approximate performance on clean data.

Ying Wang, Chiuman Ho, Wenju Xu et al.

It is critical to obtain high resolution features with long range dependency for dense prediction tasks such as semantic segmentation. To generate high-resolution output of size $H\times W$ from a low-resolution feature map of size $h\times w$ ($hw\ll HW$), a naive dense transformer incurs an intractable complexity of $\mathcal{O}(hwHW)$, limiting its application on high-resolution dense prediction. We propose a Dual-Flattening Transformer (DFlatFormer) to enable high-resolution output by reducing complexity to $\mathcal{O}(hw(H+W))$ that is multiple orders of magnitude smaller than the naive dense transformer. Decomposed queries are presented to retrieve row and column attentions tractably through separate transformers, and their outputs are combined to form a dense feature map at high resolution. To this end, the input sequence fed from an encoder is row-wise and column-wise flattened to align with decomposed queries by preserving their row and column structures, respectively. Row and column transformers also interact with each other to capture their mutual attentions with the spatial crossings between rows and columns. We also propose to perform attentions through efficient grouping and pooling to further reduce the model complexity. Extensive experiments on ADE20K and Cityscapes datasets demonstrate the superiority of the proposed dual-flattening transformer architecture with higher mIoUs.

Ziteng Cui, Yingying Zhu, Lin Gu et al.

Image restoration algorithms such as super resolution (SR) are indispensable pre-processing modules for object detection in degraded images. However, most of these algorithms assume the degradation is fixed and known a priori. When the real degradation is unknown or differs from assumption, both the pre-processing module and the consequent high-level task such as object detection would fail. Here, we propose a novel framework, RestoreDet, to detect objects in degraded low resolution images. RestoreDet utilizes the downsampling degradation as a kind of transformation for self-supervised signals to explore the equivariant representation against various resolutions and other degradation conditions. Specifically, we learn this intrinsic visual structure by encoding and decoding the degradation transformation from a pair of original and randomly degraded images. The framework could further take the advantage of advanced SR architectures with an arbitrary resolution restoring decoder to reconstruct the original correspondence from the degraded input image. Both the representation learning and object detection are optimized jointly in an end-to-end training fashion. RestoreDet is a generic framework that could be implemented on any mainstream object detection architectures. The extensive experiment shows that our framework based on CenterNet has achieved superior performance compared with existing methods when facing variant degradation situations. Our code would be released soon.

Rui Wang, Zuxuan Wu, Zejia Weng et al.

Unsupervised domain adaptation (UDA) aims to transfer knowledge learned from a fully-labeled source domain to a different unlabeled target domain. Most existing UDA methods learn domain-invariant feature representations by minimizing feature distances across domains. In this work, we build upon contrastive self-supervised learning to align features so as to reduce the domain discrepancy between training and testing sets. Exploring the same set of categories shared by both domains, we introduce a simple yet effective framework CDCL, for domain alignment. In particular, given an anchor image from one domain, we minimize its distances to cross-domain samples from the same class relative to those from different categories. Since target labels are unavailable, we use a clustering-based approach with carefully initialized centers to produce pseudo labels. In addition, we demonstrate that CDCL is a general framework and can be adapted to the data-free setting, where the source data are unavailable during training, with minimal modification. We conduct experiments on two widely used domain adaptation benchmarks, i.e., Office-31 and VisDA-2017, for image classification tasks, and demonstrate that CDCL achieves state-of-the-art performance on both datasets.

Xiang Gao, Wei Hu, Guo-Jun Qi

We present the Topology Transformation Equivariant Representation learning, a general paradigm of self-supervised learning for node representations of graph data to enable the wide applicability of Graph Convolutional Neural Networks (GCNNs). We formalize the proposed model from an information-theoretic perspective, by maximizing the mutual information between topology transformations and node representations before and after the transformations. We derive that maximizing such mutual information can be relaxed to minimizing the cross entropy between the applied topology transformation and its estimation from node representations. In particular, we seek to sample a subset of node pairs from the original graph and flip the edge connectivity between each pair to transform the graph topology. Then, we self-train a representation encoder to learn node representations by reconstructing the topology transformations from the feature representations of the original and transformed graphs. In experiments, we apply the proposed model to the downstream node classification, graph classification and link prediction tasks, and results show that the proposed method outperforms the state-of-the-art unsupervised approaches.

Guangwei Gao, Yi Yu, Jian Yang et al.

Cross-resolution face recognition (CRFR), which is important in intelligent surveillance and biometric forensics, refers to the problem of matching a low-resolution (LR) probe face image against high-resolution (HR) gallery face images. Existing shallow learning-based and deep learning-based methods focus on mapping the HR-LR face pairs into a joint feature space where the resolution discrepancy is mitigated. However, little works consider how to extract and utilize the intermediate discriminative features from the noisy LR query faces to further mitigate the resolution discrepancy due to the resolution limitations. In this study, we desire to fully exploit the multi-level deep convolutional neural network (CNN) feature set for robust CRFR. In particular, our contributions are threefold. (i) To learn more robust and discriminative features, we desire to adaptively fuse the contextual features from different layers. (ii) To fully exploit these contextual features, we design a feature set-based representation learning (FSRL) scheme to collaboratively represent the hierarchical features for more accurate recognition. Moreover, FSRL utilizes the primitive form of feature maps to keep the latent structural information, especially in noisy cases. (iii) To further promote the recognition performance, we desire to fuse the hierarchical recognition outputs from different stages. Meanwhile, the discriminability from different scales can also be fully integrated. By exploiting these advantages, the efficiency of the proposed method can be delivered. Experimental results on several face datasets have verified the superiority of the presented algorithm to the other competitive CRFR approaches.

Xiang Gao, Wei Hu, Guo-Jun Qi

3D object representation learning is a fundamental challenge in computer vision to infer about the 3D world. Recent advances in deep learning have shown their efficiency in 3D object recognition, among which view-based methods have performed best so far. However, feature learning of multiple views in existing methods is mostly performed in a supervised fashion, which often requires a large amount of data labels with high costs. In contrast, self-supervised learning aims to learn multi-view feature representations without involving labeled data. To this end, we propose a novel self-supervised paradigm to learn Multi-View Transformation Equivariant Representations (MV-TER), exploring the equivariant transformations of a 3D object and its projected multiple views. Specifically, we perform a 3D transformation on a 3D object, and obtain multiple views before and after the transformation via projection. Then, we self-train a representation to capture the intrinsic 3D object representation by decoding 3D transformation parameters from the fused feature representations of multiple views before and after the transformation. Experimental results demonstrate that the proposed MV-TER significantly outperforms the state-of-the-art view-based approaches in 3D object classification and retrieval tasks, and show the generalization to real-world datasets.