Zelin Peng

Wei Shen, Zelin Peng, Xuehui Wang et al.

The rapid development of deep learning has made a great progress in image segmentation, one of the fundamental tasks of computer vision. However, the current segmentation algorithms mostly rely on the availability of pixel-level annotations, which are often expensive, tedious, and laborious. To alleviate this burden, the past years have witnessed an increasing attention in building label-efficient, deep-learning-based image segmentation algorithms. This paper offers a comprehensive review on label-efficient image segmentation methods. To this end, we first develop a taxonomy to organize these methods according to the supervision provided by different types of weak labels (including no supervision, inexact supervision, incomplete supervision and inaccurate supervision) and supplemented by the types of segmentation problems (including semantic segmentation, instance segmentation and panoptic segmentation). Next, we summarize the existing label-efficient image segmentation methods from a unified perspective that discusses an important question: how to bridge the gap between weak supervision and dense prediction -- the current methods are mostly based on heuristic priors, such as cross-pixel similarity, cross-label constraint, cross-view consistency, and cross-image relation. Finally, we share our opinions about the future research directions for label-efficient deep image segmentation.

Zelin Peng, Guanchun Wang, Lingxi Xie et al.

Seed area generation is usually the starting point of weakly supervised semantic segmentation (WSSS). Computing the Class Activation Map (CAM) from a multi-label classification network is the de facto paradigm for seed area generation, but CAMs generated from Convolutional Neural Networks (CNNs) and Transformers are prone to be under- and over-activated, respectively, which makes the strategies to refine CAMs for CNNs usually inappropriate for Transformers, and vice versa. In this paper, we propose a Unified optimization paradigm for Seed Area GEneration (USAGE) for both types of networks, in which the objective function to be optimized consists of two terms: One is a generation loss, which controls the shape of seed areas by a temperature parameter following a deterministic principle for different types of networks; The other is a regularization loss, which ensures the consistency between the seed areas that are generated by self-adaptive network adjustment from different views, to overturn false activation in seed areas. Experimental results show that USAGE consistently improves seed area generation for both CNNs and Transformers by large margins, e.g., outperforming state-of-the-art methods by a mIoU of 4.1% on PASCAL VOC. Moreover, based on the USAGE-generated seed areas on Transformers, we achieve state-of-the-art WSSS results on both PASCAL VOC and MS COCO.

Zelin Peng, Zhengqin Xu, Zhilin Zeng et al.

Parameter-efficient fine-tuning (PEFT) is an effective methodology to unleash the potential of large foundation models in novel scenarios with limited training data. In the computer vision community, PEFT has shown effectiveness in image classification, but little research has studied its ability for image segmentation. Fine-tuning segmentation models usually require a heavier adjustment of parameters to align the proper projection directions in the parameter space for new scenarios. This raises a challenge to existing PEFT algorithms, as they often inject a limited number of individual parameters into each block, which prevents substantial adjustment of the projection direction of the parameter space due to the limitation of Hidden Markov Chain along blocks. In this paper, we equip PEFT with a cross-block orchestration mechanism to enable the adaptation of the Segment Anything Model (SAM) to various downstream scenarios. We introduce a novel inter-block communication module, which integrates a learnable relation matrix to facilitate communication among different coefficient sets of each PEFT block's parameter space. Moreover, we propose an intra-block enhancement module, which introduces a linear projection head whose weights are generated from a hyper-complex layer, further enhancing the impact of the adjustment of projection directions on the entire parameter space. Extensive experiments on diverse benchmarks demonstrate that our proposed approach consistently improves the segmentation performance significantly on novel scenarios with only around 1K additional parameters.

Guanchun Wang, Xiangrong Zhang, Zelin Peng et al.

Weakly supervised object detection (WSOD) is a challenging task, in which image-level labels (e.g., categories of the instances in the whole image) are used to train an object detector. Many existing methods follow the standard multiple instance learning (MIL) paradigm and have achieved promising performance. However, the lack of deterministic information leads to part domination and missing instances. To address these issues, this paper focuses on identifying and fully exploiting the deterministic information in WSOD. We discover that negative instances (i.e. absolutely wrong instances), ignored in most of the previous studies, normally contain valuable deterministic information. Based on this observation, we here propose a negative deterministic information (NDI) based method for improving WSOD, namely NDI-WSOD. Specifically, our method consists of two stages: NDI collecting and exploiting. In the collecting stage, we design several processes to identify and distill the NDI from negative instances online. In the exploiting stage, we utilize the extracted NDI to construct a novel negative contrastive learning mechanism and a negative guided instance selection strategy for dealing with the issues of part domination and missing instances, respectively. Experimental results on several public benchmarks including VOC 2007, VOC 2012 and MS COCO show that our method achieves satisfactory performance.

Zelin Peng, Zhengqin Xu, Zhilin Zeng et al.

Segment Anything Model (SAM) has received remarkable attention as it offers a powerful and versatile solution for object segmentation in images. However, fine-tuning SAM for downstream segmentation tasks under different scenarios remains a challenge, as the varied characteristics of different scenarios naturally requires diverse model parameter spaces. Most existing fine-tuning methods attempt to bridge the gaps among different scenarios by introducing a set of new parameters to modify SAM's original parameter space. Unlike these works, in this paper, we propose fine-tuning SAM efficiently by parameter space reconstruction (SAM-PARSER), which introduce nearly zero trainable parameters during fine-tuning. In SAM-PARSER, we assume that SAM's original parameter space is relatively complete, so that its bases are able to reconstruct the parameter space of a new scenario. We obtain the bases by matrix decomposition, and fine-tuning the coefficients to reconstruct the parameter space tailored to the new scenario by an optimal linear combination of the bases. Experimental results show that SAM-PARSER exhibits superior segmentation performance across various scenarios, while reducing the number of trainable parameters by $\approx 290$ times compared with current parameter-efficient fine-tuning methods.

Yichen Zhao, Zelin Peng, Piao Yang et al.

Radiological diagnosis is a perceptual process in which careful visual inspection and language reasoning are repeatedly interleaved. Most medical large vision language models (LVLMs) perform visual inspection only once and then rely on text-only chain-of-thought (CoT) reasoning, which operates purely in the linguistic space and is prone to hallucination. Recent methods attempt to mitigate this issue by introducing visually related coordinates, such as bounding boxes. However, these remain a pseudo-visual solution: coordinates are still text and fail to preserve rich visual details like texture and density. Motivated by the interleaved nature of radiological diagnosis, we introduce MMRad-IVL-22K, the first large-scale dataset designed for natively interleaved visual language reasoning in chest X-ray interpretation. MMRad-IVL-22K reflects a repeated cycle of reasoning and visual inspection workflow of radiologists, in which visual rationales complement textual descriptions and ground each step of the reasoning process. MMRad-IVL-22K comprises 21,994 diagnostic traces, enabling systematic scanning across 35 anatomical regions. Experimental results on advanced closed-source LVLMs demonstrate that report generation guided by multimodal CoT significantly outperforms that guided by text-only CoT in clinical accuracy and report quality (e.g., 6\% increase in the RadGraph metric), confirming that high-fidelity interleaved vision language evidence is a non-substitutable component of reliable medical AI. Furthermore, benchmarking across seven state-of-the-art open-source LVLMs demonstrates that models fine-tuned on MMRad-IVL-22K achieve superior reasoning consistency and report quality compared with both general-purpose and medical-specific LVLMs. The project page is available at https://github.com/qiuzyc/thinking_like_a_radiologist.

Guanchun Wang, Xiangrong Zhang, Zelin Peng et al.

Land cover analysis using hyperspectral images (HSI) remains an open problem due to their low spatial resolution and complex spectral information. Recent studies are primarily dedicated to designing Transformer-based architectures for spatial-spectral long-range dependencies modeling, which is computationally expensive with quadratic complexity. Selective structured state space model (Mamba), which is efficient for modeling long-range dependencies with linear complexity, has recently shown promising progress. However, its potential in hyperspectral image processing that requires handling numerous spectral bands has not yet been explored. In this paper, we innovatively propose S$^2$Mamba, a spatial-spectral state space model for hyperspectral image classification, to excavate spatial-spectral contextual features, resulting in more efficient and accurate land cover analysis. In S$^2$Mamba, two selective structured state space models through different dimensions are designed for feature extraction, one for spatial, and the other for spectral, along with a spatial-spectral mixture gate for optimal fusion. More specifically, S$^2$Mamba first captures spatial contextual relations by interacting each pixel with its adjacent through a Patch Cross Scanning module and then explores semantic information from continuous spectral bands through a Bi-directional Spectral Scanning module. Considering the distinct expertise of the two attributes in homogenous and complicated texture scenes, we realize the Spatial-spectral Mixture Gate by a group of learnable matrices, allowing for the adaptive incorporation of representations learned across different dimensions. Extensive experiments conducted on HSI classification benchmarks demonstrate the superiority and prospect of S$^2$Mamba. The code will be made available at: https://github.com/PURE-melo/S2Mamba.

Han Li, Xinyu Peng, Yaoming Wang et al.

We introduce OneCAT, a unified multimodal model that seamlessly integrates understanding, generation, and editing within a novel, pure decoder-only transformer architecture. Our framework uniquely eliminates the need for external components such as Vision Transformers (ViT) or vision tokenizer during inference, leading to significant efficiency gains, especially for high-resolution inputs. This is achieved through a modality-specific Mixture-of-Experts (MoE) structure trained with a single autoregressive (AR) objective, which also natively supports dynamic resolutions. Furthermore, we pioneer a multi-scale visual autoregressive mechanism within the Large Language Model (LLM) that drastically reduces decoding steps compared to diffusion-based methods while maintaining state-of-the-art performance. Our findings demonstrate the powerful potential of pure autoregressive modeling as a sufficient and elegant foundation for unified multimodal intelligence. As a result, OneCAT sets a new performance standard, outperforming existing open-source unified multimodal models across benchmarks for multimodal generation, editing, and understanding.

Yiheng Zhong, Zihong Luo, Chengzhi Liu et al.

Segment Anything Model (SAM) demonstrates powerful zero-shot capabilities; however, its accuracy and robustness significantly decrease when applied to medical image segmentation. Existing methods address this issue through modality fusion, integrating textual and image information to provide more detailed priors. In this study, we argue that the granularity of text and the domain gap affect the accuracy of the priors. Furthermore, the discrepancy between high-level abstract semantics and pixel-level boundary details in images can introduce noise into the fusion process. To address this, we propose Prior-Guided SAM (PG-SAM), which employs a fine-grained modality prior aligner to leverage specialized medical knowledge for better modality alignment. The core of our method lies in efficiently addressing the domain gap with fine-grained text from a medical LLM. Meanwhile, it also enhances the priors' quality after modality alignment, ensuring more accurate segmentation. In addition, our decoder enhances the model's expressive capabilities through multi-level feature fusion and iterative mask optimizer operations, supporting unprompted learning. We also propose a unified pipeline that effectively supplies high-quality semantic information to SAM. Extensive experiments on the Synapse dataset demonstrate that the proposed PG-SAM achieves state-of-the-art performance. Our code is released at https://github.com/logan-0623/PG-SAM.

Ming Hu, Kun Yuan, Yaling Shen et al.

Surgical practice involves complex visual interpretation, procedural skills, and advanced medical knowledge, making surgical vision-language pretraining (VLP) particularly challenging due to this complexity and the limited availability of annotated data. To address the gap, we propose OphCLIP, a hierarchical retrieval-augmented vision-language pretraining framework specifically designed for ophthalmic surgical workflow understanding. OphCLIP leverages the OphVL dataset we constructed, a large-scale and comprehensive collection of over 375K hierarchically structured video-text pairs with tens of thousands of different combinations of attributes (surgeries, phases/operations/actions, instruments, medications, as well as more advanced aspects like the causes of eye diseases, surgical objectives, and postoperative recovery recommendations, etc). These hierarchical video-text correspondences enable OphCLIP to learn both fine-grained and long-term visual representations by aligning short video clips with detailed narrative descriptions and full videos with structured titles, capturing intricate surgical details and high-level procedural insights, respectively. Our OphCLIP also designs a retrieval-augmented pretraining framework to leverage the underexplored large-scale silent surgical procedure videos, automatically retrieving semantically relevant content to enhance the representation learning of narrative videos. Evaluation across 11 datasets for phase recognition and multi-instrument identification shows OphCLIP's robust generalization and superior performance.

Guanchun Wang, Chenxiao Wu, Xiangrong Zhang et al.

Open-vocabulary semantic segmentation (OVSS) in remote sensing images is a promising task that employs textual descriptions for identifying undefined land cover categories. Despite notable advances, existing methods typically employ a static inference paradigm, overlooking the distinct distribution of each scene, resulting in semantic ambiguity in diverse land covers and incomplete foreground activation. Motivated by this, we propose Seeking Consensus, termed SeeCo, a plug-and-play framework to boost the performance of training-free OVSS models in remote sensing images, which recalibrates arbitrary OVSS models on-the-fly by seeking dual consensus: geometric consensus learning (GCL) through multi-view consistent observations and semantic consensus learning (SCL) via textual description adaptive calibration, which assists collaborative recalibration of visual and textual semantics. The two consensus are injected via an online consensus injector (OCI), effectively alleviating the under-activation and semantic bias. SeeCo requires no specific training process, yet recalibrates semantic-geometric alignment for each unique scene during inference. Extensive experiments on eight remote sensing OVSS benchmarks show consistent gains, proving its effectiveness and universality.

Feilong Tang, Chengzhi Liu, Zhongxing Xu et al.

Recent advancements in multimodal large language models (MLLMs) have significantly improved performance in visual question answering. However, they often suffer from hallucinations. In this work, hallucinations are categorized into two main types: initial hallucinations and snowball hallucinations. We argue that adequate contextual information can be extracted directly from the token interaction process. Inspired by causal inference in the decoding strategy, we propose to leverage causal masks to establish information propagation between multimodal tokens. The hypothesis is that insufficient interaction between those tokens may lead the model to rely on outlier tokens, overlooking dense and rich contextual cues. Therefore, we propose to intervene in the propagation process by tackling outlier tokens to enhance in-context inference. With this goal, we present FarSight, a versatile plug-and-play decoding strategy to reduce attention interference from outlier tokens merely by optimizing the causal mask. The heart of our method is effective token propagation. We design an attention register structure within the upper triangular matrix of the causal mask, dynamically allocating attention to capture attention diverted to outlier tokens. Moreover, a positional awareness encoding method with a diminishing masking rate is proposed, allowing the model to attend to further preceding tokens, especially for video sequence tasks. With extensive experiments, FarSight demonstrates significant hallucination-mitigating performance across different MLLMs on both image and video benchmarks, proving its effectiveness.

Haolin Yang, Feilong Tang, Ming Hu et al.

Video diffusion models (VDMs) facilitate the generation of high-quality videos, with current research predominantly concentrated on scaling efforts during training through improvements in data quality, computational resources, and model complexity. However, inference-time scaling has received less attention, with most approaches restricting models to a single generation attempt. Recent studies have uncovered the existence of "golden noises" that can enhance video quality during generation. Building on this, we find that guiding the scaling inference-time search of VDMs to identify better noise candidates not only evaluates the quality of the frames generated in the current step but also preserves the high-level object features by referencing the anchor frame from previous multi-chunks, thereby delivering long-term value. Our analysis reveals that diffusion models inherently possess flexible adjustments of computation by varying denoising steps, and even a one-step denoising approach, when guided by a reward signal, yields significant long-term benefits. Based on the observation, we proposeScalingNoise, a plug-and-play inference-time search strategy that identifies golden initial noises for the diffusion sampling process to improve global content consistency and visual diversity. Specifically, we perform one-step denoising to convert initial noises into a clip and subsequently evaluate its long-term value, leveraging a reward model anchored by previously generated content. Moreover, to preserve diversity, we sample candidates from a tilted noise distribution that up-weights promising noises. In this way, ScalingNoise significantly reduces noise-induced errors, ensuring more coherent and spatiotemporally consistent video generation. Extensive experiments on benchmark datasets demonstrate that the proposed ScalingNoise effectively improves long video generation.

Yu Huang, Zelin Peng, Yichen Zhao et al.

Medical image segmentation is crucial for clinical diagnosis, yet existing models are limited by their reliance on explicit human instructions and lack the active reasoning capabilities to understand complex clinical questions. While recent advancements in multimodal large language models (MLLMs) have improved medical question-answering (QA) tasks, most methods struggle to generate precise segmentation masks, limiting their application in automatic medical diagnosis. In this paper, we introduce medical image reasoning segmentation, a novel task that aims to generate segmentation masks based on complex and implicit medical instructions. To address this, we propose MedSeg-R, an end-to-end framework that leverages the reasoning abilities of MLLMs to interpret clinical questions while also capable of producing corresponding precise segmentation masks for medical images. It is built on two core components: 1) a global context understanding module that interprets images and comprehends complex medical instructions to generate multi-modal intermediate tokens, and 2) a pixel-level grounding module that decodes these tokens to produce precise segmentation masks and textual responses. Furthermore, we introduce MedSeg-QA, a large-scale dataset tailored for the medical image reasoning segmentation task. It includes over 10,000 image-mask pairs and multi-turn conversations, automatically annotated using large language models and refined through physician reviews. Experiments show MedSeg-R's superior performance across several benchmarks, achieving high segmentation accuracy and enabling interpretable textual analysis of medical images.

Guanchun Wang, Xiangrong Zhang, Yifei Zhang et al.

Unsupervised anomaly detection in hyperspectral images (HSI), aiming to detect unknown targets from backgrounds, is challenging for earth surface monitoring. However, current studies are hindered by steep computational costs due to the high-dimensional property of HSI and dense sampling-based training paradigm, constraining their rapid deployment. Our key observation is that, during training, not all samples within the same homogeneous area are indispensable, whereas ingenious sampling can provide a powerful substitute for reducing costs. Motivated by this, we propose an Asymmetrical Consensus State Space Model (ACMamba) to significantly reduce computational costs without compromising accuracy. Specifically, we design an asymmetrical anomaly detection paradigm that utilizes region-level instances as an efficient alternative to dense pixel-level samples. In this paradigm, a low-cost Mamba-based module is introduced to discover global contextual attributes of regions that are essential for HSI reconstruction. Additionally, we develop a consensus learning strategy from the optimization perspective to simultaneously facilitate background reconstruction and anomaly compression, further alleviating the negative impact of anomaly reconstruction. Theoretical analysis and extensive experiments across eight benchmarks verify the superiority of ACMamba, demonstrating a faster speed and stronger performance over the state-of-the-art.

Zelin Peng, Zhengqin Xu, Qingyang Liu et al.

Multi-modal large language models (MLLMs) have emerged as a transformative approach for aligning visual and textual understanding. They typically require extremely high computational resources (e.g., thousands of GPUs) for training to achieve cross-modal alignment at multi-granularity levels. We argue that a key source of this inefficiency lies in the vision encoders they widely equip with, e.g., CLIP and SAM, which lack the alignment with language at multi-granularity levels. To address this issue, in this paper, we leverage hyperbolic space, which inherently models hierarchical levels and thus provides a principled framework for bridging the granularity gap between visual and textual modalities at an arbitrary granularity level. Concretely, we propose an efficient training paradigm for MLLMs, dubbed as HyperET, which can optimize visual representations to align with their textual counterparts at an arbitrary granularity level through dynamic hyperbolic radius adjustment in hyperbolic space. HyperET employs learnable matrices with Möbius multiplication operations, implemented via three effective configurations: diagonal scaling matrices, block-diagonal matrices, and banded matrices, providing a flexible yet efficient parametrization strategy. Comprehensive experiments across multiple MLLM benchmarks demonstrate that HyperET consistently improves both existing pre-training and fine-tuning MLLMs clearly with less than 1\% additional parameters.

Yu Huang, Zelin Peng, Changsong Wen et al.

Affordance segmentation aims to parse 3D objects into functionally distinct parts, bridging recognition and interaction for applications in robotic manipulation, embodied AI, and AR. While recent studies leverage visual or textual prompts to guide this process, they often rely on point cloud encoders as generic feature extractors, overlooking the intrinsic challenges of 3D data such as sparsity, noise, and geometric ambiguity. As a result, 3D features learned in isolation frequently lack clear and semantically consistent functional boundaries. To address this bottleneck, we propose a semantic-grounded learning paradigm that transfers rich semantic knowledge from large-scale 2D Vision Foundation Models (VFMs) into the 3D domain. Specifically, We introduce Cross-Modal Affinity Transfer (CMAT), a pre-training strategy that aligns a 3D encoder with lifted 2D semantics and jointly optimizes reconstruction, affinity, and diversity to yield semantically organized representations. Building on this backbone, we further design the Cross-modal Affordance Segmentation Transformer (CAST), which integrates multi-modal prompts with CMAT-pretrained features to generate precise, prompt-aware segmentation maps. Extensive experiments on standard benchmarks demonstrate that our framework establishes new state-of-the-art results for 3D affordance segmentation.

Qingyang Liu, Bingjie Gao, Weiheng Huang et al.

3D scene reconstruction and 4D human animation have seen rapid progress and broad adoption in recent years. However, seamlessly integrating reconstructed scenes with 4D human animation to produce visually engaging results remains challenging. One key difficulty lies in placing the human at the correct location and scale within the scene while avoiding unrealistic interpenetration. Another challenge is that the human and the background may exhibit different lighting and style, leading to unrealistic composites. In addition, appealing character motion videos are often accompanied by camera movements, which means that the viewpoints need to be reconstructed along a specified trajectory. We present AnimateScene, which addresses the above issues in a unified framework. First, we design an accurate placement module that automatically determines a plausible 3D position for the human and prevents any interpenetration within the scene during motion. Second, we propose a training-free style alignment method that adapts the 4D human representation to match the background's lighting and style, achieving coherent visual integration. Finally, we design a joint post-reconstruction method for both the 4D human and the 3D scene that allows camera trajectories to be inserted, enabling the final rendered video to feature visually appealing camera movements. Extensive experiments show that AnimateScene generates dynamic scene videos with high geometric detail and spatiotemporal coherence across various camera and action combinations.

Zelin Peng, Yichen Zhao, Yu Huang et al.

Computer-aided medical image analysis is crucial for disease diagnosis and treatment planning, yet limited annotated datasets restrict medical-specific model development. While vision-language models (VLMs) like CLIP offer strong generalization capabilities, their direct application to medical imaging analysis is impeded by a significant domain gap. Existing approaches to bridge this gap, including prompt learning and one-way modality interaction techniques, typically focus on introducing domain knowledge to a single modality. Although this may offer performance gains, it often causes modality misalignment, thereby failing to unlock the full potential of VLMs. In this paper, we propose \textbf{NEARL-CLIP} (i\underline{N}teracted qu\underline{E}ry \underline{A}daptation with o\underline{R}thogona\underline{L} Regularization), a novel cross-modality interaction VLM-based framework that contains two contributions: (1) Unified Synergy Embedding Transformer (USEformer), which dynamically generates cross-modality queries to promote interaction between modalities, thus fostering the mutual enrichment and enhancement of multi-modal medical domain knowledge; (2) Orthogonal Cross-Attention Adapter (OCA). OCA introduces an orthogonality technique to decouple the new knowledge from USEformer into two distinct components: the truly novel information and the incremental knowledge. By isolating the learning process from the interference of incremental knowledge, OCA enables a more focused acquisition of new information, thereby further facilitating modality interaction and unleashing the capability of VLMs. Notably, NEARL-CLIP achieves these two contributions in a parameter-efficient style, which only introduces \textbf{1.46M} learnable parameters.

Xiangrong Zhang, Zelin Peng, Peng Zhu et al.

Semantic segmentation has been continuously investigated in the last ten years, and majority of the established technologies are based on supervised models. In recent years, image-level weakly supervised semantic segmentation (WSSS), including single- and multi-stage process, has attracted large attention due to data labeling efficiency. In this paper, we propose to embed affinity learning of multi-stage approaches in a single-stage model. To be specific, we introduce an adaptive affinity loss to thoroughly learn the local pairwise affinity. As such, a deep neural network is used to deliver comprehensive semantic information in the training phase, whilst improving the performance of the final prediction module. On the other hand, considering the existence of errors in the pseudo labels, we propose a novel label reassign loss to mitigate over-fitting. Extensive experiments are conducted on the PASCAL VOC 2012 dataset to evaluate the effectiveness of our proposed approach that outperforms other standard single-stage methods and achieves comparable performance against several multi-stage methods.