Qilong Wang

Jiangtao Xie, Fei Long, Jiaming Lv et al.

Few-shot classification is a challenging problem as only very few training examples are given for each new task. One of the effective research lines to address this challenge focuses on learning deep representations driven by a similarity measure between a query image and few support images of some class. Statistically, this amounts to measure the dependency of image features, viewed as random vectors in a high-dimensional embedding space. Previous methods either only use marginal distributions without considering joint distributions, suffering from limited representation capability, or are computationally expensive though harnessing joint distributions. In this paper, we propose a deep Brownian Distance Covariance (DeepBDC) method for few-shot classification. The central idea of DeepBDC is to learn image representations by measuring the discrepancy between joint characteristic functions of embedded features and product of the marginals. As the BDC metric is decoupled, we formulate it as a highly modular and efficient layer. Furthermore, we instantiate DeepBDC in two different few-shot classification frameworks. We make experiments on six standard few-shot image benchmarks, covering general object recognition, fine-grained categorization and cross-domain classification. Extensive evaluations show our DeepBDC significantly outperforms the counterparts, while establishing new state-of-the-art results. The source code is available at http://www.peihuali.org/DeepBDC

Wei Shang, Dongwei Ren, Wanying Zhang et al.

Modern consumer cameras commonly employ the rolling shutter (RS) imaging mechanism, via which images are captured by scanning scenes row-by-row, resulting in RS distortion for dynamic scenes. To correct RS distortion, existing methods adopt a fully supervised learning manner that requires high framerate global shutter (GS) images as ground-truth for supervision. In this paper, we propose an enhanced Self-supervised learning framework for Dual reversed RS distortion Correction (SelfDRSC++). Firstly, we introduce a lightweight DRSC network that incorporates a bidirectional correlation matching block to refine the joint optimization of optical flows and corrected RS features, thereby improving correction performance while reducing network parameters. Subsequently, to effectively train the DRSC network, we propose a self-supervised learning strategy that ensures cycle consistency between input and reconstructed dual reversed RS images. The RS reconstruction in SelfDRSC++ can be interestingly formulated as a specialized instance of video frame interpolation, where each row in reconstructed RS images is interpolated from predicted GS images by utilizing RS distortion time maps. By achieving superior performance while simplifying the training process, SelfDRSC++ enables feasible one-stage self-supervised training. Additionally, besides start and end RS scanning time, SelfDRSC++ allows supervision of GS images at arbitrary intermediate scanning times, thus enabling the learned DRSC network to generate high framerate GS videos. The code and trained models are available at \url{https://github.com/shangwei5/SelfDRSC_plusplus}.

Mingze Gao, Qilong Wang, Zhenyi Lin et al.

Recently, efficient fine-tuning of large-scale pre-trained models has attracted increasing research interests, where linear probing (LP) as a fundamental module is involved in exploiting the final representations for task-dependent classification. However, most of the existing methods focus on how to effectively introduce a few of learnable parameters, and little work pays attention to the commonly used LP module. In this paper, we propose a novel Moment Probing (MP) method to further explore the potential of LP. Distinguished from LP which builds a linear classification head based on the mean of final features (e.g., word tokens for ViT) or classification tokens, our MP performs a linear classifier on feature distribution, which provides the stronger representation ability by exploiting richer statistical information inherent in features. Specifically, we represent feature distribution by its characteristic function, which is efficiently approximated by using first- and second-order moments of features. Furthermore, we propose a multi-head convolutional cross-covariance (MHC$^3$) to compute second-order moments in an efficient and effective manner. By considering that MP could affect feature learning, we introduce a partially shared module to learn two recalibrating parameters (PSRP) for backbones based on MP, namely MP$_{+}$. Extensive experiments on ten benchmarks using various models show that our MP significantly outperforms LP and is competitive with counterparts at less training cost, while our MP$_{+}$ achieves state-of-the-art performance.

Qilong Wang, Xiaofan Ming, Zhenyi Lin et al.

Virtual furniture synthesis, which seamlessly integrates reference objects into indoor scenes while maintaining geometric coherence and visual realism, holds substantial promise for home design and e-commerce applications. However, this field remains underexplored due to the scarcity of reproducible benchmarks and the limitations of existing image composition methods in achieving high-fidelity furniture synthesis while preserving background integrity. To overcome these challenges, we first present RoomBench++, a comprehensive and publicly available benchmark dataset tailored for this task. It consists of 112,851 training pairs and 1,832 testing pairs drawn from both real-world indoor videos and realistic home design renderings, thereby supporting robust training and evaluation under practical conditions. Then, we propose RoomEditor++, a versatile diffusion-based architecture featuring a parameter-sharing dual diffusion backbone, which is compatible with both U-Net and DiT architectures. This design unifies the feature extraction and inpainting processes for reference and background images. Our in-depth analysis reveals that the parameter-sharing mechanism enforces aligned feature representations, facilitating precise geometric transformations, texture preservation, and seamless integration. Extensive experiments validate that RoomEditor++ is superior over state-of-the-art approaches in terms of quantitative metrics, qualitative assessments, and human preference studies, while highlighting its strong generalization to unseen indoor scenes and general scenes without task-specific fine-tuning. The dataset and source code are available at \url{https://github.com/stonecutter-21/roomeditor}.

Shaobo Wang, Yantai Yang, Qilong Wang et al.

Dataset Distillation (DD) aims to synthesize a small dataset capable of performing comparably to the original dataset. Despite the success of numerous DD methods, theoretical exploration of this area remains unaddressed. In this paper, we take an initial step towards understanding various matching-based DD methods from the perspective of sample difficulty. We begin by empirically examining sample difficulty, measured by gradient norm, and observe that different matching-based methods roughly correspond to specific difficulty tendencies. We then extend the neural scaling laws of data pruning to DD to theoretically explain these matching-based methods. Our findings suggest that prioritizing the synthesis of easier samples from the original dataset can enhance the quality of distilled datasets, especially in low IPC (image-per-class) settings. Based on our empirical observations and theoretical analysis, we introduce the Sample Difficulty Correction (SDC) approach, designed to predominantly generate easier samples to achieve higher dataset quality. Our SDC can be seamlessly integrated into existing methods as a plugin with minimal code adjustments. Experimental results demonstrate that adding SDC generates higher-quality distilled datasets across 7 distillation methods and 6 datasets.

Shihao Zhu, Ziheng Ouyang, Yijia Kang et al.

Diffusion-based stylization has advanced significantly, yet existing methods are limited to color-driven transformations, neglecting complex semantics and material details.We introduce StyleExpert, a semantic-aware framework based on the Mixture of Experts (MoE). Our framework employs a unified style encoder, trained on our large-scale dataset of content-style-stylized triplets, to embed diverse styles into a consistent latent space. This embedding is then used to condition a similarity-aware gating mechanism, which dynamically routes styles to specialized experts within the MoE architecture. Leveraging this MoE architecture, our method adeptly handles diverse styles spanning multiple semantic levels, from shallow textures to deep semantics. Extensive experiments show that StyleExpert outperforms existing approaches in preserving semantics and material details, while generalizing to unseen styles. Our code and collected images are available at the project page: https://hh-lg.github.io/StyleExpert-Page/.

Liwei Luo, Shuaitengyuan Li, Dongwei Ren et al.

Recently, remarkable progress has been made in large-scale pre-trained model tuning, and inference efficiency is becoming more crucial for practical deployment. Early exiting in conjunction with multi-stage predictors, when cooperated with a parameter-efficient fine-tuning strategy, offers a straightforward way to achieve an inference-efficient model. However, a key challenge remains unresolved: How can early stages provide low-level fundamental features to deep stages while simultaneously supplying high-level discriminative features to early-stage predictors? To address this problem, we propose a Decoupled Multi-Predictor Optimization (DMPO) method to effectively decouple the low-level representative ability and high-level discriminative ability in early stages. First, in terms of architecture, we introduce a lightweight bypass module into multi-stage predictors for functional decomposition of shallow features from early stages, while a high-order statistics-based predictor is developed for early stages to effectively enhance their discriminative ability. To reasonably train our multi-predictor architecture, a decoupled optimization is proposed to allocate two-phase loss weights for multi-stage predictors during model tuning, where the initial training phase enables the model to prioritize the acquisition of discriminative ability of deep stages via emphasizing representative ability of early stages, and the latter training phase drives discriminative ability towards earlier stages as much as possible. As such, our DMPO can effectively decouple representative and discriminative abilities in early stages in terms of architecture design and model optimization. Experiments across various datasets and pre-trained backbones demonstrate that DMPO clearly outperforms its counterparts when reducing computational cost.

Yu Wang, Junxian Mu, Hongzhi Huang et al.

Open set recognition (OSR) requires models to classify known samples while detecting unknown samples for real-world applications. Existing studies show impressive progress using unknown samples from auxiliary datasets to regularize OSR models, but they have proved to be sensitive to selecting such known outliers. In this paper, we discuss the aforementioned problem from a new perspective: Can we regularize OSR models without elaborately selecting auxiliary known outliers? We first empirically and theoretically explore the role of foregrounds and backgrounds in open set recognition and disclose that: 1) backgrounds that correlate with foregrounds would mislead the model and cause failures when encounters 'partially' known images; 2) Backgrounds unrelated to foregrounds can serve as auxiliary known outliers and provide regularization via global average pooling. Based on the above insights, we propose a new method, Background Mix (BackMix), that mixes the foreground of an image with different backgrounds to remove the underlying fore-background priors. Specifically, BackMix first estimates the foreground with class activation maps (CAMs), then randomly replaces image patches with backgrounds from other images to obtain mixed images for training. With backgrounds de-correlated from foregrounds, the open set recognition performance is significantly improved. The proposed method is quite simple to implement, requires no extra operation for inferences, and can be seamlessly integrated into almost all of the existing frameworks. The code is released on https://github.com/Vanixxz/BackMix.

Zixian Guo, Ming Liu, Qilong Wang et al.

Current large vision-language models (LVLMs) typically employ a connector module to link visual features with text embeddings of large language models (LLMs) and use end-to-end training to achieve multi-modal understanding in a unified process. Effective alignment needs high-quality pre-training data and a carefully designed training process. Current LVLMs face challenges when addressing complex vision-language reasoning tasks, with their reasoning capabilities notably lagging behind those of LLMs. This paper proposes a paradigm shift: instead of training end-to-end vision-language reasoning models, we advocate for developing a decoupled reasoning framework based on existing visual interpretation specialists and text-based reasoning LLMs. Our approach leverages (1) a dedicated vision-language model to transform the visual content of images into textual descriptions and (2) an LLM to perform reasoning according to the visual-derived text and the original question. This method presents a cost-efficient solution for multi-modal model development by optimizing existing models to work collaboratively, avoiding end-to-end development of vision-language models from scratch. By transforming images into language model-compatible text representations, it facilitates future low-cost and flexible upgrades to upcoming powerful LLMs. We introduce an outcome-rewarded joint-tuning strategy to optimize the cooperation between the visual interpretation and linguistic reasoning model. Evaluation results on vision-language benchmarks demonstrate that the decoupled reasoning framework outperforms recent LVLMs. Our approach yields particularly significant performance gains on visually intensive geometric mathematics problems. The code is available: https://github.com/guozix/DVLR.

Bowen Dong, Zitong Huang, Yuelin Guo et al.

Weakly-supervised object detection (WSOD) has emerged as an inspiring recent topic to avoid expensive instance-level object annotations. However, the bounding boxes of most existing WSOD methods are mainly determined by precomputed proposals, thereby being limited in precise object localization. In this paper, we defend the problem setting for improving localization performance by leveraging the bounding box regression knowledge from a well-annotated auxiliary dataset. First, we use the well-annotated auxiliary dataset to explore a series of learnable bounding box adjusters (LBBAs) in a multi-stage training manner, which is class-agnostic. Then, only LBBAs and a weakly-annotated dataset with non-overlapped classes are used for training LBBA-boosted WSOD. As such, our LBBAs are practically more convenient and economical to implement while avoiding the leakage of the auxiliary well-annotated dataset. In particular, we formulate learning bounding box adjusters as a bi-level optimization problem and suggest an EM-like multi-stage training algorithm. Then, a multi-stage scheme is further presented for LBBA-boosted WSOD. Additionally, a masking strategy is adopted to improve proposal classification. Experimental results verify the effectiveness of our method. Our method performs favorably against state-of-the-art WSOD methods and knowledge transfer model with similar problem setting. Code is publicly available at \url{https://github.com/DongSky/lbba_boosted_wsod}.

Dongwei Ren, Kai Zhang, Qilong Wang et al.

Blind deconvolution is a classical yet challenging low-level vision problem with many real-world applications. Traditional maximum a posterior (MAP) based methods rely heavily on fixed and handcrafted priors that certainly are insufficient in characterizing clean images and blur kernels, and usually adopt specially designed alternating minimization to avoid trivial solution. In contrast, existing deep motion deblurring networks learn from massive training images the mapping to clean image or blur kernel, but are limited in handling various complex and large size blur kernels. To connect MAP and deep models, we in this paper present two generative networks for respectively modeling the deep priors of clean image and blur kernel, and propose an unconstrained neural optimization solution to blind deconvolution. In particular, we adopt an asymmetric Autoencoder with skip connections for generating latent clean image, and a fully-connected network (FCN) for generating blur kernel. Moreover, the SoftMax nonlinearity is applied to the output layer of FCN to meet the non-negative and equality constraints. The process of neural optimization can be explained as a kind of "zero-shot" self-supervised learning of the generative networks, and thus our proposed method is dubbed SelfDeblur. Experimental results show that our SelfDeblur can achieve notable quantitative gains as well as more visually plausible deblurring results in comparison to state-of-the-art blind deconvolution methods on benchmark datasets and real-world blurry images. The source code is available at https://github.com/csdwren/SelfDeblur

Hao Wang, Qilong Wang, Fan Yang et al.

Collection of massive well-annotated samples is effective in improving object detection performance but is extremely laborious and costly. Instead of data collection and annotation, the recently proposed Cut-Paste methods [12, 15] show the potential to augment training dataset by cutting foreground objects and pasting them on proper new backgrounds. However, existing Cut-Paste methods cannot guarantee synthetic images always precisely model visual context, and all of them require external datasets. To handle above issues, this paper proposes a simple yet effective instance-switching (IS) strategy, which generates new training data by switching instances of same class from different images. Our IS naturally preserves contextual coherence in the original images while requiring no external dataset. For guiding our IS to obtain better object performance, we explore issues of instance imbalance and class importance in datasets, which frequently occur and bring adverse effect on detection performance. To this end, we propose a novel Progressive and Selective Instance-Switching (PSIS) method to augment training data for object detection. The proposed PSIS enhances instance balance by combining selective re-sampling with a class-balanced loss, and considers class importance by progressively augmenting training dataset guided by detection performance. The experiments are conducted on the challenging MS COCO benchmark, and results demonstrate our PSIS brings clear improvement over various state-of-the-art detectors (e.g., Faster R-CNN, FPN, Mask R-CNN and SNIPER), showing the superiority and generality of our PSIS. Code and models are available at: https://github.com/Hwang64/PSIS.

Hao Wang, Qilong Wang, Mingqi Gao et al.

Although Faster R-CNN and its variants have shown promising performance in object detection, they only exploit simple first-order representation of object proposals for final classification and regression. Recent classification methods demonstrate that the integration of high-order statistics into deep convolutional neural networks can achieve impressive improvement, but their goal is to model whole images by discarding location information so that they cannot be directly adopted to object detection. In this paper, we make an attempt to exploit high-order statistics in object detection, aiming at generating more discriminative representations for proposals to enhance the performance of detectors. To this end, we propose a novel Multi-scale Location-aware Kernel Representation (MLKP) to capture high-order statistics of deep features in proposals. Our MLKP can be efficiently computed on a modified multi-scale feature map using a low-dimensional polynomial kernel approximation.Moreover, different from existing orderless global representations based on high-order statistics, our proposed MLKP is location retentive and sensitive so that it can be flexibly adopted to object detection. Through integrating into Faster R-CNN schema, the proposed MLKP achieves very competitive performance with state-of-the-art methods, and improves Faster R-CNN by 4.9% (mAP), 4.7% (mAP) and 5.0% (AP at IOU=[0.5:0.05:0.95]) on PASCAL VOC 2007, VOC 2012 and MS COCO benchmarks, respectively. Code is available at: https://github.com/Hwang64/MLKP.

Yuwei Tang, Zhenyi Lin, Qilong Wang et al.

Recently, pre-trained vision-language models (e.g., CLIP) have shown great potential in few-shot learning and attracted a lot of research interest. Although efforts have been made to improve few-shot ability of CLIP, key factors on the effectiveness of existing methods have not been well studied, limiting further exploration of CLIP's potential in few-shot learning. In this paper, we first introduce a unified formulation to analyze CLIP-based few-shot learning methods from a perspective of logit bias, which encourages us to learn an effective logit bias for further improving performance of CLIP-based few-shot learning methods. To this end, we disassemble three key components involved in computation of logit bias (i.e., logit features, logit predictor, and logit fusion) and empirically analyze the effect on performance of few-shot classification. Based on analysis of key components, this paper proposes a novel AMU-Tuning method to learn effective logit bias for CLIP-based few-shot classification. Specifically, our AMU-Tuning predicts logit bias by exploiting the appropriate $\underline{\textbf{A}}$uxiliary features, which are fed into an efficient feature-initialized linear classifier with $\underline{\textbf{M}}$ulti-branch training. Finally, an $\underline{\textbf{U}}$ncertainty-based fusion is developed to incorporate logit bias into CLIP for few-shot classification. The experiments are conducted on several widely used benchmarks, and the results show AMU-Tuning clearly outperforms its counterparts while achieving state-of-the-art performance of CLIP-based few-shot learning without bells and whistles.

Tianyi Zhu, Dongwei Ren, Qilong Wang et al.

Generative inbetweening aims to generate intermediate frame sequences by utilizing two key frames as input. Although remarkable progress has been made in video generation models, generative inbetweening still faces challenges in maintaining temporal stability due to the ambiguous interpolation path between two key frames. This issue becomes particularly severe when there is a large motion gap between input frames. In this paper, we propose a straightforward yet highly effective Frame-wise Conditions-driven Video Generation (FCVG) method that significantly enhances the temporal stability of interpolated video frames. Specifically, our FCVG provides an explicit condition for each frame, making it much easier to identify the interpolation path between two input frames and thus ensuring temporally stable production of visually plausible video frames. To achieve this, we suggest extracting matched lines from two input frames that can then be easily interpolated frame by frame, serving as frame-wise conditions seamlessly integrated into existing video generation models. In extensive evaluations covering diverse scenarios such as natural landscapes, complex human poses, camera movements and animations, existing methods often exhibit incoherent transitions across frames. In contrast, our FCVG demonstrates the capability to generate temporally stable videos using both linear and non-linear interpolation curves. Our project page and code are available at \url{https://fcvg-inbetween.github.io/}.

Bing Cao, Xingxin Xu, Pengfei Zhu et al.

Image fusion aims to integrate complementary information from multiple input images acquired through various sources to synthesize a new fused image. Existing methods usually employ distinct constraint designs tailored to specific scenes, forming fixed fusion paradigms. However, this data-driven fusion approach is challenging to deploy in varying scenarios, especially in rapidly changing environments. To address this issue, we propose a conditional controllable fusion (CCF) framework for general image fusion tasks without specific training. Due to the dynamic differences of different samples, our CCF employs specific fusion constraints for each individual in practice. Given the powerful generative capabilities of the denoising diffusion model, we first inject the specific constraints into the pre-trained DDPM as adaptive fusion conditions. The appropriate conditions are dynamically selected to ensure the fusion process remains responsive to the specific requirements in each reverse diffusion stage. Thus, CCF enables conditionally calibrating the fused images step by step. Extensive experiments validate our effectiveness in general fusion tasks across diverse scenarios against the competing methods without additional training.

Yilong Wang, Zilin Gao, Qilong Wang et al.

Going beyond few-shot action recognition (FSAR), cross-domain FSAR (CDFSAR) has attracted recent research interests by solving the domain gap lying in source-to-target transfer learning. Existing CDFSAR methods mainly focus on joint training of source and target data to mitigate the side effect of domain gap. However, such kind of methods suffer from two limitations: First, pair-wise joint training requires retraining deep models in case of one source data and multiple target ones, which incurs heavy computation cost, especially for large source and small target data. Second, pre-trained models after joint training are adopted to target domain in a straightforward manner, hardly taking full potential of pre-trained models and then limiting recognition performance. To overcome above limitations, this paper proposes a simple yet effective baseline, namely Temporal-Aware Model Tuning (TAMT) for CDFSAR. Specifically, our TAMT involves a decoupled paradigm by performing pre-training on source data and fine-tuning target data, which avoids retraining for multiple target data with single source. To effectively and efficiently explore the potential of pre-trained models in transferring to target domain, our TAMT proposes a Hierarchical Temporal Tuning Network (HTTN), whose core involves local temporal-aware adapters (TAA) and a global temporal-aware moment tuning (GTMT). Particularly, TAA learns few parameters to recalibrate the intermediate features of frozen pre-trained models, enabling efficient adaptation to target domains. Furthermore, GTMT helps to generate powerful video representations, improving match performance on the target domain. Experiments on several widely used video benchmarks show our TAMT outperforms the recently proposed counterparts by 13%$\sim$31%, achieving new state-of-the-art CDFSAR results.

Bing Cao, Quanhao Lu, Jiekang Feng et al.

The dynamic imbalance of the fore-background is a major challenge in video object counting, which is usually caused by the sparsity of target objects. This remains understudied in existing works and often leads to severe under-/over-prediction errors. To tackle this issue in video object counting, we propose a density-embedded Efficient Masked Autoencoder Counting (E-MAC) framework in this paper. To empower the model's representation ability on density regression, we develop a new $\mathtt{D}$ensity-$\mathtt{E}$mbedded $\mathtt{M}$asked m$\mathtt{O}$deling ($\mathtt{DEMO}$) method, which first takes the density map as an auxiliary modality to perform multimodal self-representation learning for image and density map. Although $\mathtt{DEMO}$ contributes to effective cross-modal regression guidance, it also brings in redundant background information, making it difficult to focus on the foreground regions. To handle this dilemma, we propose an efficient spatial adaptive masking derived from density maps to boost efficiency. Meanwhile, we employ an optical flow-based temporal collaborative fusion strategy to effectively capture the dynamic variations across frames, aligning features to derive multi-frame density residuals. The counting accuracy of the current frame is boosted by harnessing the information from adjacent frames. In addition, considering that most existing datasets are limited to human-centric scenarios, we first propose a large video bird counting dataset, DroneBird, in natural scenarios for migratory bird protection. Extensive experiments on three crowd datasets and our \textit{DroneBird} validate our superiority against the counterparts. The code and dataset are available.

Quan-Sheng Zeng, Yunheng Li, Qilong Wang et al.

Visual token compression is critical for Large Vision-Language Models (LVLMs) to efficiently process high-resolution inputs. Existing methods that typically adopt fixed compression ratios cannot adapt to scenes of varying complexity, often causing imprecise pruning that discards informative visual tokens and results in degraded model performance. To address this issue, we introduce a dynamic pruning framework, GlimpsePrune, inspired by human cognition. It takes a data-driven ''glimpse'' and prunes irrelevant visual tokens in a single forward pass before answer generation. This approach prunes 92.6% of visual tokens while on average fully retaining the baseline performance on free-form VQA tasks. The reduced computational cost also enables more effective fine-tuning: an enhanced GlimpsePrune+ achieves 110% of the baseline performance while maintaining a similarly high pruning rate. Our work paves a new way for building more powerful and efficient LVLMs.

Wenlong Yu, Qilong Wang, Chuang Liu et al.

Explainability is a critical factor influencing the wide deployment of deep vision models (DVMs). Concept-based post-hoc explanation methods can provide both global and local insights into model decisions. However, current methods in this field face challenges in that they are inflexible to automatically construct accurate and sufficient linguistic explanations for global concepts and local circuits. Particularly, the intrinsic polysemanticity in semantic Visual Concepts (VCs) impedes the interpretability of concepts and DVMs, which is underestimated severely. In this paper, we propose a Chain-of-Explanation (CoE) approach to address these issues. Specifically, CoE automates the decoding and description of VCs to construct global concept explanation datasets. Further, to alleviate the effect of polysemanticity on model explainability, we design a concept polysemanticity disentanglement and filtering mechanism to distinguish the most contextually relevant concept atoms. Besides, a Concept Polysemanticity Entropy (CPE), as a measure of model interpretability, is formulated to quantify the degree of concept uncertainty. The modeling of deterministic concepts is upgraded to uncertain concept atom distributions. Finally, CoE automatically enables linguistic local explanations of the decision-making process of DVMs by tracing the concept circuit. GPT-4o and human-based experiments demonstrate the effectiveness of CPE and the superiority of CoE, achieving an average absolute improvement of 36% in terms of explainability scores.

Xiaojie Yin, Qilong Wang, Bing Cao et al.

Recently, many studies have been conducted to enhance the zero-shot generalization ability of vision-language models (e.g., CLIP) by addressing the semantic misalignment between image and text embeddings in downstream tasks. Although many efforts have been made, existing methods barely consider the fact that a class of images can be described by notably different textual concepts due to well-known lexical variation in natural language processing, which heavily affects the zero-shot generalization of CLIP. Therefore, this paper proposes a \textbf{S}ynonymous \textbf{S}emantic \textbf{S}pace ($S^3$) for each image class, rather than relying on a single textual concept, achieving more stable semantic alignment and improving the zero-shot generalization of CLIP. Specifically, our $S^3$ method first generates several synonymous concepts based on the label of each class by using large language models, and constructs a continuous yet compact synonymous semantic space based on the Vietoris-Rips complex of the generated synonymous concepts. Furthermore, we explore the effect of several point-to-space metrics on our $S^3$, while presenting a point-to-local-center metric to compute similarity between image embeddings and the synonymous semantic space of each class, accomplishing effective zero-shot predictions. Extensive experiments are conducted across 17 benchmarks, including fine-grained zero-shot classification, natural distribution zero-shot classification, and open-vocabulary segmentation, and the results show that our $S^3$ outperforms state-of-the-art methods.

Zilin Gao, Qilong Wang, Bingbing Zhang et al.

Thanks to capability to alleviate the cost of large-scale annotation, few-shot action recognition (FSAR) has attracted increased attention of researchers in recent years. Existing FSAR approaches typically neglect the role of individual motion pattern in comparison, and under-explore the feature statistics for video dynamics. Thereby, they struggle to handle the challenging temporal misalignment in video dynamics, particularly by using 2D backbones. To overcome these limitations, this work proposes an adaptively aligned multi-scale second-order moment network, namely A$^2$M$^2$-Net, to describe the latent video dynamics with a collection of powerful representation candidates and adaptively align them in an instance-guided manner. To this end, our A$^2$M$^2$-Net involves two core components, namely, adaptive alignment (A$^2$ module) for matching, and multi-scale second-order moment (M$^2$ block) for strong representation. Specifically, M$^2$ block develops a collection of semantic second-order descriptors at multiple spatio-temporal scales. Furthermore, A$^2$ module aims to adaptively select informative candidate descriptors while considering the individual motion pattern. By such means, our A$^2$M$^2$-Net is able to handle the challenging temporal misalignment problem by establishing an adaptive alignment protocol for strong representation. Notably, our proposed method generalizes well to various few-shot settings and diverse metrics. The experiments are conducted on five widely used FSAR benchmarks, and the results show our A$^2$M$^2$-Net achieves very competitive performance compared to state-of-the-arts, demonstrating its effectiveness and generalization.

Xiaojie Yin, Qilong Wang, Qinghua Hu

Vision-language models (VLMs) pre-trained on web-scale data exhibit promising zero-shot generalization but often suffer from semantic misalignment due to domain gaps between pre-training and downstream tasks. Existing approaches primarily focus on text prompting with class-specific descriptions and visual-text adaptation via aligning cropped image regions with textual descriptions. However, they still face the issues of incomplete textual prompts and noisy visual prompts. In this paper, we propose a novel constrained prompt enhancement (CPE) method to improve visual-textual alignment by constructing comprehensive textual prompts and compact visual prompts from the semantic perspective. Specifically, our approach consists of two key components: Topology-Guided Synonymous Semantic Generation (TGSSG) and Category-Agnostic Discriminative Region Selection (CADRS). Textually, to address the issue of incomplete semantic expression in textual prompts, our TGSSG first generates synonymous semantic set for each category via large language models, and constructs comprehensive textual prompts based on semantic ambiguity entropy and persistent homology analysis. Visually, to mitigate the irrelevant visual noise introduced by random cropping, our CADRS identifies discriminative regions with activation maps outputted by a pre-trained vision model, effectively filtering out noisy regions and generating compact visual prompts. Given the comprehensive set of textual prompts and compact set of visual prompts, we introduce two set-to-set matching strategies based on test-time adaptation (TTA) and optimal transport (OT) to achieve effective visual-textual alignment, and so improve zero-shot generalization of VLMs.

Junjie Wu, Jiangtao Xie, Zhaolin Zhang et al.

Recently, Contrastive Language-Image Pre-training (CLIP) has shown promising performance in domain-specific data (e.g., biology), and has attracted increasing research attention. Existing works generally focus on collecting extensive domain-specific data and directly tuning the original CLIP models. Intuitively, such a paradigm takes no full consideration of the characteristics lying in domain-specific data (e.g., fine-grained nature of biological data) and so limits model capability, while mostly losing the original ability of CLIP in the general domain. In this paper, we propose a Distribution Alignment-based Language-Image Pre-Training (DALIP) method for biological data. Specifically, DALIP optimizes CLIP models by matching the similarity between feature distribution of image-text pairs instead of the original [cls] token, which can capture rich yet effective information inherent in image-text pairs as powerful representations, and so better cope with fine-grained nature of biological data. Particularly, our DALIP efficiently approximates feature distribution via its first- and second-order statistics, while presenting a Multi-head Brownian Distance Covariance (MBDC) module to acquire second-order statistics of token features efficiently. Furthermore, we collect a new dataset for plant domain (e.g., specific data in biological domain) comprising 10M plant data with 3M general-domain data (namely PlantMix-13M) according to data mixing laws. Extensive experiments show that DALIP clearly outperforms existing CLIP counterparts in biological domain, while well generalizing to remote sensing and medical imaging domains. Besides, our PlantMix-13M dataset further boosts performance of DALIP in plant domain, while preserving model ability in general domain.

Yuanze Li, Chun-Mei Feng, Qilong Wang et al.

Human beings can leverage knowledge from relative tasks to improve learning on a primary task. Similarly, multi-task learning methods suggest using auxiliary tasks to enhance a neural network's performance on a specific primary task. However, previous methods often select auxiliary tasks carefully but treat them as secondary during training. The weights assigned to auxiliary losses are typically smaller than the primary loss weight, leading to insufficient training on auxiliary tasks and ultimately failing to support the main task effectively. To address this issue, we propose an uncertainty-based impartial learning method that ensures balanced training across all tasks. Additionally, we consider both gradients and uncertainty information during backpropagation to further improve performance on the primary task. Extensive experiments show that our method achieves performance comparable to or better than state-of-the-art approaches. Moreover, our weighting strategy is effective and robust in enhancing the performance of the primary task regardless the noise auxiliary tasks' pseudo labels.

Wenlong Yu, Dongyue Chen, Qilong Wang et al.

Fine-grained domain generalization (FGDG) is a more challenging task than traditional DG tasks due to its small inter-class variations and relatively large intra-class disparities. When domain distribution changes, the vulnerability of subtle features leads to a severe deterioration in model performance. Nevertheless, humans inherently demonstrate the capacity for generalizing to out-of-distribution data, leveraging structured multi-granularity knowledge that emerges from discerning the commonality and specificity within categories. Likewise, we propose a Feature Structuralized Domain Generalization (FSDG) model, wherein features experience structuralization into common, specific, and confounding segments, harmoniously aligned with their relevant semantic concepts, to elevate performance in FGDG. Specifically, feature structuralization (FS) is accomplished through joint optimization of five constraints: a decorrelation function applied to disentangled segments, three constraints ensuring common feature consistency and specific feature distinctiveness, and a prediction calibration term. By imposing these stipulations, FSDG is prompted to disentangle and align features based on multi-granularity knowledge, facilitating robust subtle distinctions among categories. Extensive experimentation on three benchmarks consistently validates the superiority of FSDG over state-of-the-art counterparts, with an average improvement of 6.2% in FGDG performance. Beyond that, the explainability analysis on explicit concept matching intensity between the shared concepts among categories and the model channels, along with experiments on various mainstream model architectures, substantiates the validity of FS.

Yunheng Li, Zhongyu Li, Shanghua Gao et al.

Effectively modeling discriminative spatio-temporal information is essential for segmenting activities in long action sequences. However, we observe that existing methods are limited in weak spatio-temporal modeling capability due to two forms of decoupled modeling: (i) cascaded interaction couples spatial and temporal modeling, which over-smooths motion modeling over the long sequence, and (ii) joint-shared temporal modeling adopts shared weights to model each joint, ignoring the distinct motion patterns of different joints. We propose a Decoupled Spatio-Temporal Framework (DeST) to address the above issues. Firstly, we decouple the cascaded spatio-temporal interaction to avoid stacking multiple spatio-temporal blocks, while achieving sufficient spatio-temporal interaction. Specifically, DeST performs once unified spatial modeling and divides the spatial features into different groups of subfeatures, which then adaptively interact with temporal features from different layers. Since the different sub-features contain distinct spatial semantics, the model could learn the optimal interaction pattern at each layer. Meanwhile, inspired by the fact that different joints move at different speeds, we propose joint-decoupled temporal modeling, which employs independent trainable weights to capture distinctive temporal features of each joint. On four large-scale benchmarks of different scenes, DeST significantly outperforms current state-of-the-art methods with less computational complexity.

Zilin Gao, Qilong Wang, Bingbing Zhang et al.

For video recognition task, a global representation summarizing the whole contents of the video snippets plays an important role for the final performance. However, existing video architectures usually generate it by using a simple, global average pooling (GAP) method, which has limited ability to capture complex dynamics of videos. For image recognition task, there exist evidences showing that covariance pooling has stronger representation ability than GAP. Unfortunately, such plain covariance pooling used in image recognition is an orderless representative, which cannot model spatio-temporal structure inherent in videos. Therefore, this paper proposes a Temporal-attentive Covariance Pooling(TCP), inserted at the end of deep architectures, to produce powerful video representations. Specifically, our TCP first develops a temporal attention module to adaptively calibrate spatio-temporal features for the succeeding covariance pooling, approximatively producing attentive covariance representations. Then, a temporal covariance pooling performs temporal pooling of the attentive covariance representations to characterize both intra-frame correlations and inter-frame cross-correlations of the calibrated features. As such, the proposed TCP can capture complex temporal dynamics. Finally, a fast matrix power normalization is introduced to exploit geometry of covariance representations. Note that our TCP is model-agnostic and can be flexibly integrated into any video architectures, resulting in TCPNet for effective video recognition. The extensive experiments on six benchmarks (e.g., Kinetics, Something-Something V1 and Charades) using various video architectures show our TCPNet is clearly superior to its counterparts, while having strong generalization ability. The source code is publicly available.

Longyin Wen, Dawei Du, Pengfei Zhu et al.

To promote the developments of object detection, tracking and counting algorithms in drone-captured videos, we construct a benchmark with a new drone-captured largescale dataset, named as DroneCrowd, formed by 112 video clips with 33,600 HD frames in various scenarios. Notably, we annotate 20,800 people trajectories with 4.8 million heads and several video-level attributes. Meanwhile, we design the Space-Time Neighbor-Aware Network (STNNet) as a strong baseline to solve object detection, tracking and counting jointly in dense crowds. STNNet is formed by the feature extraction module, followed by the density map estimation heads, and localization and association subnets. To exploit the context information of neighboring objects, we design the neighboring context loss to guide the association subnet training, which enforces consistent relative position of nearby objects in temporal domain. Extensive experiments on our DroneCrowd dataset demonstrate that STNNet performs favorably against the state-of-the-arts.

Jiangtao Xie, Ruiren Zeng, Qilong Wang et al.

Transformer models are not only successful in natural language processing (NLP) but also demonstrate high potential in computer vision (CV). Despite great advance, most of works only focus on improvement of architectures but pay little attention to the classification head. For years transformer models base exclusively on classification token to construct the final classifier, without explicitly harnessing high-level word tokens. In this paper, we propose a novel transformer model called second-order transformer (SoT), exploiting simultaneously the classification token and word tokens for the classifier. Specifically, we empirically disclose that high-level word tokens contain rich information, which per se are very competent with the classifier and moreover, are complementary to the classification token. To effectively harness such rich information, we propose multi-headed global cross-covariance pooling with singular value power normalization, which shares similar philosophy and thus is compatible with the transformer block, better than commonly used pooling methods. Then, we study comprehensively how to explicitly combine word tokens with classification token for building the final classification head. For CV tasks, our SoT significantly improves state-of-the-art vision transformers on challenging benchmarks including ImageNet and ImageNet-A. For NLP tasks, through fine-tuning based on pretrained language transformers including GPT and BERT, our SoT greatly boosts the performance on widely used tasks such as CoLA and RTE. Code will be available at https://peihuali.org/SoT

Qilong Wang, Li Zhang, Banggu Wu et al.

Recent works have demonstrated that global covariance pooling (GCP) has the ability to improve performance of deep convolutional neural networks (CNNs) on visual classification task. Despite considerable advance, the reasons on effectiveness of GCP on deep CNNs have not been well studied. In this paper, we make an attempt to understand what deep CNNs benefit from GCP in a viewpoint of optimization. Specifically, we explore the effect of GCP on deep CNNs in terms of the Lipschitzness of optimization loss and the predictiveness of gradients, and show that GCP can make the optimization landscape more smooth and the gradients more predictive. Furthermore, we discuss the connection between GCP and second-order optimization for deep CNNs. More importantly, above findings can account for several merits of covariance pooling for training deep CNNs that have not been recognized previously or fully explored, including significant acceleration of network convergence (i.e., the networks trained with GCP can support rapid decay of learning rates, achieving favorable performance while significantly reducing number of training epochs), stronger robustness to distorted examples generated by image corruptions and perturbations, and good generalization ability to different vision tasks, e.g., object detection and instance segmentation. We conduct extensive experiments using various deep CNN models on diversified tasks, and the results provide strong support to our findings.

Longyin Wen, Dawei Du, Pengfei Zhu et al.

This paper proposes a space-time multi-scale attention network (STANet) to solve density map estimation, localization and tracking in dense crowds of video clips captured by drones with arbitrary crowd density, perspective, and flight altitude. Our STANet method aggregates multi-scale feature maps in sequential frames to exploit the temporal coherency, and then predict the density maps, localize the targets, and associate them in crowds simultaneously. A coarse-to-fine process is designed to gradually apply the attention module on the aggregated multi-scale feature maps to enforce the network to exploit the discriminative space-time features for better performance. The whole network is trained in an end-to-end manner with the multi-task loss, formed by three terms, i.e., the density map loss, localization loss and association loss. The non-maximal suppression followed by the min-cost flow framework is used to generate the trajectories of targets' in scenarios. Since existing crowd counting datasets merely focus on crowd counting in static cameras rather than density map estimation, counting and tracking in crowds on drones, we have collected a new large-scale drone-based dataset, DroneCrowd, formed by 112 video clips with 33,600 high resolution frames (i.e., 1920x1080) captured in 70 different scenarios. With intensive amount of effort, our dataset provides 20,800 people trajectories with 4.8 million head annotations and several video-level attributes in sequences. Extensive experiments are conducted on two challenging public datasets, i.e., Shanghaitech and UCF-QNRF, and our DroneCrowd, to demonstrate that STANet achieves favorable performance against the state-of-the-arts. The datasets and codes can be found at https://github.com/VisDrone.

Qilong Wang, Banggu Wu, Pengfei Zhu et al.

Recently, channel attention mechanism has demonstrated to offer great potential in improving the performance of deep convolutional neural networks (CNNs). However, most existing methods dedicate to developing more sophisticated attention modules for achieving better performance, which inevitably increase model complexity. To overcome the paradox of performance and complexity trade-off, this paper proposes an Efficient Channel Attention (ECA) module, which only involves a handful of parameters while bringing clear performance gain. By dissecting the channel attention module in SENet, we empirically show avoiding dimensionality reduction is important for learning channel attention, and appropriate cross-channel interaction can preserve performance while significantly decreasing model complexity. Therefore, we propose a local cross-channel interaction strategy without dimensionality reduction, which can be efficiently implemented via $1D$ convolution. Furthermore, we develop a method to adaptively select kernel size of $1D$ convolution, determining coverage of local cross-channel interaction. The proposed ECA module is efficient yet effective, e.g., the parameters and computations of our modules against backbone of ResNet50 are 80 vs. 24.37M and 4.7e-4 GFLOPs vs. 3.86 GFLOPs, respectively, and the performance boost is more than 2% in terms of Top-1 accuracy. We extensively evaluate our ECA module on image classification, object detection and instance segmentation with backbones of ResNets and MobileNetV2. The experimental results show our module is more efficient while performing favorably against its counterparts.

Qilong Wang, Jiangtao Xie, Wangmeng Zuo et al.

Compared with global average pooling in existing deep convolutional neural networks (CNNs), global covariance pooling can capture richer statistics of deep features, having potential for improving representation and generalization abilities of deep CNNs. However, integration of global covariance pooling into deep CNNs brings two challenges: (1) robust covariance estimation given deep features of high dimension and small sample size; (2) appropriate usage of geometry of covariances. To address these challenges, we propose a global Matrix Power Normalized COVariance (MPN-COV) Pooling. Our MPN-COV conforms to a robust covariance estimator, very suitable for scenario of high dimension and small sample size. It can also be regarded as Power-Euclidean metric between covariances, effectively exploiting their geometry. Furthermore, a global Gaussian embedding network is proposed to incorporate first-order statistics into MPN-COV. For fast training of MPN-COV networks, we implement an iterative matrix square root normalization, avoiding GPU unfriendly eigen-decomposition inherent in MPN-COV. Additionally, progressive 1x1 convolutions and group convolution are introduced to compress covariance representations. The proposed methods are highly modular, readily plugged into existing deep CNNs. Extensive experiments are conducted on large-scale object classification, scene categorization, fine-grained visual recognition and texture classification, showing our methods outperform the counterparts and obtain state-of-the-art performance.

Zilin Gao, Jiangtao Xie, Qilong Wang et al.

Deep Convolutional Networks (ConvNets) are fundamental to, besides large-scale visual recognition, a lot of vision tasks. As the primary goal of the ConvNets is to characterize complex boundaries of thousands of classes in a high-dimensional space, it is critical to learn higher-order representations for enhancing non-linear modeling capability. Recently, Global Second-order Pooling (GSoP), plugged at the end of networks, has attracted increasing attentions, achieving much better performance than classical, first-order networks in a variety of vision tasks. However, how to effectively introduce higher-order representation in earlier layers for improving non-linear capability of ConvNets is still an open problem. In this paper, we propose a novel network model introducing GSoP across from lower to higher layers for exploiting holistic image information throughout a network. Given an input 3D tensor outputted by some previous convolutional layer, we perform GSoP to obtain a covariance matrix which, after nonlinear transformation, is used for tensor scaling along channel dimension. Similarly, we can perform GSoP along spatial dimension for tensor scaling as well. In this way, we can make full use of the second-order statistics of the holistic image throughout a network. The proposed networks are thoroughly evaluated on large-scale ImageNet-1K, and experiments have shown that they outperformed non-trivially the counterparts while achieving state-of-the-art results.

Peihua Li, Jiangtao Xie, Qilong Wang et al.

Global covariance pooling in convolutional neural networks has achieved impressive improvement over the classical first-order pooling. Recent works have shown matrix square root normalization plays a central role in achieving state-of-the-art performance. However, existing methods depend heavily on eigendecomposition (EIG) or singular value decomposition (SVD), suffering from inefficient training due to limited support of EIG and SVD on GPU. Towards addressing this problem, we propose an iterative matrix square root normalization method for fast end-to-end training of global covariance pooling networks. At the core of our method is a meta-layer designed with loop-embedded directed graph structure. The meta-layer consists of three consecutive nonlinear structured layers, which perform pre-normalization, coupled matrix iteration and post-compensation, respectively. Our method is much faster than EIG or SVD based ones, since it involves only matrix multiplications, suitable for parallel implementation on GPU. Moreover, the proposed network with ResNet architecture can converge in much less epochs, further accelerating network training. On large-scale ImageNet, we achieve competitive performance superior to existing counterparts. By finetuning our models pre-trained on ImageNet, we establish state-of-the-art results on three challenging fine-grained benchmarks. The source code and network models will be available at http://www.peihuali.org/iSQRT-COV

Hongliang Yan, Yukang Ding, Peihua Li et al.

In domain adaptation, maximum mean discrepancy (MMD) has been widely adopted as a discrepancy metric between the distributions of source and target domains. However, existing MMD-based domain adaptation methods generally ignore the changes of class prior distributions, i.e., class weight bias across domains. This remains an open problem but ubiquitous for domain adaptation, which can be caused by changes in sample selection criteria and application scenarios. We show that MMD cannot account for class weight bias and results in degraded domain adaptation performance. To address this issue, a weighted MMD model is proposed in this paper. Specifically, we introduce class-specific auxiliary weights into the original MMD for exploiting the class prior probability on source and target domains, whose challenge lies in the fact that the class label in target domain is unavailable. To account for it, our proposed weighted MMD model is defined by introducing an auxiliary weight for each class in the source domain, and a classification EM algorithm is suggested by alternating between assigning the pseudo-labels, estimating auxiliary weights and updating model parameters. Extensive experiments demonstrate the superiority of our weighted MMD over conventional MMD for domain adaptation.

Peihua Li, Jiangtao Xie, Qilong Wang et al.

By stacking layers of convolution and nonlinearity, convolutional networks (ConvNets) effectively learn from low-level to high-level features and discriminative representations. Since the end goal of large-scale recognition is to delineate complex boundaries of thousands of classes, adequate exploration of feature distributions is important for realizing full potentials of ConvNets. However, state-of-the-art works concentrate only on deeper or wider architecture design, while rarely exploring feature statistics higher than first-order. We take a step towards addressing this problem. Our method consists in covariance pooling, instead of the most commonly used first-order pooling, of high-level convolutional features. The main challenges involved are robust covariance estimation given a small sample of large-dimensional features and usage of the manifold structure of covariance matrices. To address these challenges, we present a Matrix Power Normalized Covariance (MPN-COV) method. We develop forward and backward propagation formulas regarding the nonlinear matrix functions such that MPN-COV can be trained end-to-end. In addition, we analyze both qualitatively and quantitatively its advantage over the well-known Log-Euclidean metric. On the ImageNet 2012 validation set, by combining MPN-COV we achieve over 4%, 3% and 2.5% gains for AlexNet, VGG-M and VGG-16, respectively; integration of MPN-COV into 50-layer ResNet outperforms ResNet-101 and is comparable to ResNet-152. The source code will be available on the project page: http://www.peihuali.org/MPN-COV

Qilong Wang, Peihua Li, Lei Zhang et al.

The bag-of-features (BoF) model for image classification has been thoroughly studied over the last decade. Different from the widely used BoF methods which modeled images with a pre-trained codebook, the alternative codebook free image modeling method, which we call Codebookless Model (CLM), attracted little attention. In this paper, we present an effective CLM that represents an image with a single Gaussian for classification. By embedding Gaussian manifold into a vector space, we show that the simple incorporation of our CLM into a linear classifier achieves very competitive accuracy compared with state-of-the-art BoF methods (e.g., Fisher Vector). Since our CLM lies in a high dimensional Riemannian manifold, we further propose a joint learning method of low-rank transformation with support vector machine (SVM) classifier on the Gaussian manifold, in order to reduce computational and storage cost. To study and alleviate the side effect of background clutter on our CLM, we also present a simple yet effective partial background removal method based on saliency detection. Experiments are extensively conducted on eight widely used databases to demonstrate the effectiveness and efficiency of our CLM method.