Hebei Li

Hebei Li, Yueyi Zhang, Zhiwei Xiong et al.

U-Net, known for its simple yet efficient architecture, is widely utilized for image processing tasks and is particularly suitable for deployment on neuromorphic chips. This paper introduces the novel concept of Spiking-UNet for image processing, which combines the power of Spiking Neural Networks (SNNs) with the U-Net architecture. To achieve an efficient Spiking-UNet, we face two primary challenges: ensuring high-fidelity information propagation through the network via spikes and formulating an effective training strategy. To address the issue of information loss, we introduce multi-threshold spiking neurons, which improve the efficiency of information transmission within the Spiking-UNet. For the training strategy, we adopt a conversion and fine-tuning pipeline that leverage pre-trained U-Net models. During the conversion process, significant variability in data distribution across different parts is observed when utilizing skip connections. Therefore, we propose a connection-wise normalization method to prevent inaccurate firing rates. Furthermore, we adopt a flow-based training method to fine-tune the converted models, reducing time steps while preserving performance. Experimental results show that, on image segmentation and denoising, our Spiking-UNet achieves comparable performance to its non-spiking counterpart, surpassing existing SNN methods. Compared with the converted Spiking-UNet without fine-tuning, our Spiking-UNet reduces inference time by approximately 90\%. This research broadens the application scope of SNNs in image processing and is expected to inspire further exploration in the field of neuromorphic engineering. The code for our Spiking-UNet implementation is available at https://github.com/SNNresearch/Spiking-UNet.

Yansong Peng, Hebei Li, Peixi Wu et al.

We introduce D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR models. D-FINE comprises two key components: Fine-grained Distribution Refinement (FDR) and Global Optimal Localization Self-Distillation (GO-LSD). FDR transforms the regression process from predicting fixed coordinates to iteratively refining probability distributions, providing a fine-grained intermediate representation that significantly enhances localization accuracy. GO-LSD is a bidirectional optimization strategy that transfers localization knowledge from refined distributions to shallower layers through self-distillation, while also simplifying the residual prediction tasks for deeper layers. Additionally, D-FINE incorporates lightweight optimizations in computationally intensive modules and operations, achieving a better balance between speed and accuracy. Specifically, D-FINE-L / X achieves 54.0% / 55.8% AP on the COCO dataset at 124 / 78 FPS on an NVIDIA T4 GPU. When pretrained on Objects365, D-FINE-L / X attains 57.1% / 59.3% AP, surpassing all existing real-time detectors. Furthermore, our method significantly enhances the performance of a wide range of DETR models by up to 5.3% AP with negligible extra parameters and training costs. Our code and pretrained models: https://github.com/Peterande/D-FINE.

Feipeng Ma, Yizhou Zhou, Zheyu Zhang et al.

Recent advancements in Multimodal Large Language Models (MLLMs) have demonstrated satisfactory performance across various vision-language tasks. Current approaches for vision and language interaction fall into two categories: self-attention-based and cross-attention-based methods. However, both approaches present inherent limitations, forcing a trade-off between data and computational efficiency. To address this issue, we introduce the Data-$\textbf{E}$fficient and Compute-$\textbf{E}$fficient $\textbf{MLLM}$ ($\textbf{EE-MLLM}$). Specifically, we modify the original self-attention mechanism in MLLM to a composite attention mechanism. This mechanism has two key characteristics: 1) eliminating the computational overhead of self-attention among visual tokens to achieve $\textbf{compute efficiency}$, and 2) reusing the weights from each layer of LLM to facilitate effective vision-language modality alignment for $\textbf{data efficiency}$. As a result, EE-MLLM significantly outperforms Flamingo with limited training data, and reduces the prefilling time to 79 ms on an H800 GPU, compared to LLaVA's 277 ms. To further investigate the efficiency of EE-MLLM, we present a training-free variant named EE-MLLM-F, which reduces the computation cost of self-attention-based method without additional training. Experimental results demonstrate the effectiveness of EE-MLLM across a range of benchmarks, including general-purpose datasets like MMBench and SeedBench, as well as fine-grained tasks such as TextVQA and DocVQA.

Yansong Peng, Hebei Li, Yueyi Zhang et al.

While recent Transformer-based approaches have shown impressive performances on event-based object detection tasks, their high computational costs still diminish the low power consumption advantage of event cameras. Image-based works attempt to reduce these costs by introducing sparse Transformers. However, they display inadequate sparsity and adaptability when applied to event-based object detection, since these approaches cannot balance the fine granularity of token-level sparsification and the efficiency of window-based Transformers, leading to reduced performance and efficiency. Furthermore, they lack scene-specific sparsity optimization, resulting in information loss and a lower recall rate. To overcome these limitations, we propose the Scene Adaptive Sparse Transformer (SAST). SAST enables window-token co-sparsification, significantly enhancing fault tolerance and reducing computational overhead. Leveraging the innovative scoring and selection modules, along with the Masked Sparse Window Self-Attention, SAST showcases remarkable scene-aware adaptability: It focuses only on important objects and dynamically optimizes sparsity level according to scene complexity, maintaining a remarkable balance between performance and computational cost. The evaluation results show that SAST outperforms all other dense and sparse networks in both performance and efficiency on two large-scale event-based object detection datasets (1Mpx and Gen1). Code: https://github.com/Peterande/SAST

Zhe Qian, Nianbing Su, Zhonghua Wang et al.

Current multimodal models often suffer from shallow reasoning, leading to errors caused by incomplete or inconsistent thought processes. To address this limitation, we propose Self-Verification and Self-Rectification (SVSR), a unified framework that explicitly integrates self-verification and self-rectification into the model's reasoning pipeline, substantially improving robustness and reliability in complex visual understanding and multimodal reasoning tasks. SVSR is built on a novel three-stage training paradigm. First, we construct a high-quality unified preference dataset by refining reasoning traces from pre-trained vision-language models, incorporating both forward and backward reasoning to embed self-reflective signals. Second, we perform cold-start supervised fine-tuning on this dataset to learn structured, multi-step reasoning behaviors. Third, we apply a Semi-online Direct Preference Optimization (Semi-online DPO) process, continuously augmenting the training corpus with high-quality, model-generated reasoning traces filtered by a powerful teacher VLM. This pipeline enables the model to learn, elicit, and refine its ability to self-verify and self-rectify. Extensive experiments across diverse benchmarks demonstrate that SVSR improves reasoning accuracy and enables stronger generalization to unseen tasks and question types. Notably, once trained with explicit self-reflective reasoning, the model also exhibits improved implicit reasoning ability, outperforming strong baselines even when no explicit reasoning traces are provided. These results highlight the potential of SVSR for building more dependable, introspective, and cognitively aligned multimodal systems.

Peixi Wu, Ke Mei, Feipeng Ma et al.

Multimodal Large Language Models (MLLMs) have emerged as a promising foundation for universal multimodal embeddings. Recent studies have shown that reasoning-driven generative multimodal embeddings can outperform discriminative embeddings on several embedding tasks. However, Chain-of-Thought (CoT) reasoning tends to generate redundant thinking steps and introduce semantic ambiguity in the summarized answers in broader retrieval scenarios. To address this limitation, we propose Rewrite-driven Multimodal Embedding (RIME), a unified framework that jointly optimizes generation and embedding through a retrieval-friendly rewrite. Meanwhile, we present the Cross-Mode Alignment (CMA) to bridge the generative and discriminative embedding spaces, enabling flexible mutual retrieval to trade off efficiency and accuracy. Based on this, we also introduce Refine Reinforcement Learning (Refine-RL) that treats discriminative embeddings as stable semantic anchors to guide the rewrite optimization. Extensive experiments on MMEB-V2, MRMR and UVRB demonstrate that RIME substantially outperforms prior generative embedding models while significantly reducing the length of thinking.

Yansong Peng, Kai Zhu, Yu Liu et al.

Continuous-time Consistency Models (CMs) promise efficient few-step generation but face significant challenges with training instability. We argue this instability stems from a fundamental conflict: by training a network to learn only a shortcut across a probability flow, the model loses its grasp on the instantaneous velocity field that defines the flow. Our solution is to explicitly anchor the model in the underlying flow during training. We introduce the Flow-Anchored Consistency Model (FACM), a simple but effective training strategy that uses a Flow Matching (FM) task as an anchor for the primary CM shortcut objective. This Flow-Anchoring approach requires no architectural modifications and is broadly compatible with standard model architectures. By distilling a pre-trained LightningDiT model, our method achieves a state-of-the-art FID of 1.32 with two steps (NFE=2) and 1.76 with just one step (NFE=1) on ImageNet 256x256, significantly outperforming previous methods. This provides a general and effective recipe for building high-performance, few-step generative models. Our code and pretrained models: https://github.com/ali-vilab/FACM.

Zhangchi Hu, Yifan Zhao, Yansong Peng et al.

We present RiO-DETR: DETR for Real-time Oriented Object Detection, the first real-time oriented detection transformer to the best of our knowledge. Adapting DETR to oriented bounding boxes (OBBs) poses three challenges: semantics-dependent orientation, angle periodicity that breaks standard Euclidean refinement, and an enlarged search space that slows convergence. RiO-DETR resolves these issues with task-native designs while preserving real-time efficiency. First, we propose Content-Driven Angle Estimation by decoupling angle from positional queries, together with Rotation-Rectified Orthogonal Attention to capture complementary cues for reliable orientation. Second, Decoupled Periodic Refinement combines bounded coarse-to-fine updates with a Shortest-Path Periodic Loss for stable learning across angular seams. Third, Oriented Dense O2O injects angular diversity into dense supervision to speed up angle convergence at no extra cost. Extensive experiments on DOTA-1.0, DIOR-R, and FAIR-1M-2.0 demonstrate RiO-DETR establishes a new speed--accuracy trade-off for real-time oriented detection. Code will be made publicly available.

Jie Chen, Zhangchi Hu, Peixi Wu et al.

Dynamic scene reconstruction is a long-term challenge in 3D vision. Existing plane-based methods in dynamic Gaussian splatting suffer from an unsuitable low-rank assumption, causing feature overlap and poor rendering quality. Although 4D hash encoding provides an explicit representation without low-rank constraints, directly applying it to the entire dynamic scene leads to substantial hash collisions and redundancy. To address these challenges, we present DASH, a real-time dynamic scene rendering framework that employs 4D hash encoding coupled with self-supervised decomposition. Our approach begins with a self-supervised decomposition mechanism that separates dynamic and static components without manual annotations or precomputed masks. Next, we introduce a multiresolution 4D hash encoder for dynamic elements, providing an explicit representation that avoids the low-rank assumption. Finally, we present a spatio-temporal smoothness regularization strategy to mitigate unstable deformation artifacts. Experiments on real-world datasets demonstrate that DASH achieves state-of-the-art dynamic rendering performance, exhibiting enhanced visual quality at real-time speeds of 264 FPS on a single 4090 GPU. Code: https://github.com/chenj02/DASH.

Zhangchi Hu, Peixi Wu, Jie Chen et al.

Tiny object detection plays a vital role in drone surveillance, remote sensing, and autonomous systems, enabling the identification of small targets across vast landscapes. However, existing methods suffer from inefficient feature leverage and high computational costs due to redundant feature processing and rigid query allocation. To address these challenges, we propose Dome-DETR, a novel framework with Density-Oriented Feature-Query Manipulation for Efficient Tiny Object Detection. To reduce feature redundancies, we introduce a lightweight Density-Focal Extractor (DeFE) to produce clustered compact foreground masks. Leveraging these masks, we incorporate Masked Window Attention Sparsification (MWAS) to focus computational resources on the most informative regions via sparse attention. Besides, we propose Progressive Adaptive Query Initialization (PAQI), which adaptively modulates query density across spatial areas for better query allocation. Extensive experiments demonstrate that Dome-DETR achieves state-of-the-art performance (+3.3 AP on AI-TOD-V2 and +2.5 AP on VisDrone) while maintaining low computational complexity and a compact model size. Code is available at https://github.com/RicePasteM/Dome-DETR.

Fanding Huang, Jingyan Jiang, Qinting Jiang et al.

Recent vision-language models (VLMs) face significant challenges in test-time adaptation to novel domains. While cache-based methods show promise by leveraging historical information, they struggle with both caching unreliable feature-label pairs and indiscriminately using single-class information during querying, significantly compromising adaptation accuracy. To address these limitations, we propose COSMIC (Clique-Oriented Semantic Multi-space Integration for CLIP), a robust test-time adaptation framework that enhances adaptability through multi-granular, cross-modal semantic caching and graph-based querying mechanisms. Our framework introduces two key innovations: Dual Semantics Graph (DSG) and Clique Guided Hyper-class (CGH). The Dual Semantics Graph constructs complementary semantic spaces by incorporating textual features, coarse-grained CLIP features, and fine-grained DINOv2 features to capture rich semantic relationships. Building upon these dual graphs, the Clique Guided Hyper-class component leverages structured class relationships to enhance prediction robustness through correlated class selection. Extensive experiments demonstrate COSMIC's superior performance across multiple benchmarks, achieving significant improvements over state-of-the-art methods: 15.81% gain on out-of-distribution tasks and 5.33% on cross-domain generation with CLIP RN-50. Code is available at github.com/hf618/COSMIC.

Hebei Li, Jin Wang, Jiahui Yuan et al.

In the realm of video object segmentation (VOS), the challenge of operating under low-light conditions persists, resulting in notably degraded image quality and compromised accuracy when comparing query and memory frames for similarity computation. Event cameras, characterized by their high dynamic range and ability to capture motion information of objects, offer promise in enhancing object visibility and aiding VOS methods under such low-light conditions. This paper introduces a pioneering framework tailored for low-light VOS, leveraging event camera data to elevate segmentation accuracy. Our approach hinges on two pivotal components: the Adaptive Cross-Modal Fusion (ACMF) module, aimed at extracting pertinent features while fusing image and event modalities to mitigate noise interference, and the Event-Guided Memory Matching (EGMM) module, designed to rectify the issue of inaccurate matching prevalent in low-light settings. Additionally, we present the creation of a synthetic LLE-DAVIS dataset and the curation of a real-world LLE-VOS dataset, encompassing frames and events. Experimental evaluations corroborate the efficacy of our method across both datasets, affirming its effectiveness in low-light scenarios.

Yuqing Wen, Hebei Li, Kefan Gu et al.

The rapid progress of auto-regressive vision-language models (VLMs) has inspired growing interest in vision-language-action models (VLA) for robotic manipulation. Recently, masked diffusion models, a paradigm distinct from autoregressive models, have begun to demonstrate competitive performance in text generation and multimodal applications, leading to the development of a series of diffusion-based VLMs (d-VLMs). However, leveraging such models for robot policy learning remains largely unexplored. In this work, we present LLaDA-VLA, the first Vision-Language-Diffusion-Action model built upon pretrained d-VLMs for robotic manipulation. To effectively adapt d-VLMs to robotic domain, we introduce two key designs: (1) a localized special-token classification strategy that replaces full-vocabulary classification with special action token classification, reducing adaptation difficulty; (2) a hierarchical action-structured decoding strategy that decodes action sequences hierarchically considering the dependencies within and across actions. Extensive experiments demonstrate that LLaDA-VLA significantly outperforms state-of-the-art VLAs on both simulation and real-world robots.

Peixi Wu, Bosong Chai, Hebei Li et al.

Spiking Neural Networks (SNNs) offer an attractive and energy-efficient alternative to conventional Artificial Neural Networks (ANNs) due to their sparse binary activation. When SNN meets Transformer, it shows great potential in 2D image processing. However, their application for 3D point cloud remains underexplored. To this end, we present Spiking Point Transformer (SPT), the first transformer-based SNN framework for point cloud classification. Specifically, we first design Queue-Driven Sampling Direct Encoding for point cloud to reduce computational costs while retaining the most effective support points at each time step. We introduce the Hybrid Dynamics Integrate-and-Fire Neuron (HD-IF), designed to simulate selective neuron activation and reduce over-reliance on specific artificial neurons. SPT attains state-of-the-art results on three benchmark datasets that span both real-world and synthetic datasets in the SNN domain. Meanwhile, the theoretical energy consumption of SPT is at least 6.4$\times$ less than its ANN counterpart.

Hebei Li, Yansong Peng, Jiahui Yuan et al.

Event cameras have recently been introduced into image semantic segmentation, owing to their high temporal resolution and other advantageous properties. However, existing event-based semantic segmentation methods often fail to fully exploit the complementary information provided by frames and events, resulting in complex training strategies and increased computational costs. To address these challenges, we propose an efficient hybrid framework for image semantic segmentation, comprising a Spiking Neural Network branch for events and an Artificial Neural Network branch for frames. Specifically, we introduce three specialized modules to facilitate the interaction between these two branches: the Adaptive Temporal Weighting (ATW) Injector, the Event-Driven Sparse (EDS) Injector, and the Channel Selection Fusion (CSF) module. The ATW Injector dynamically integrates temporal features from event data into frame features, enhancing segmentation accuracy by leveraging critical dynamic temporal information. The EDS Injector effectively combines sparse event data with rich frame features, ensuring precise temporal and spatial information alignment. The CSF module selectively merges these features to optimize segmentation performance. Experimental results demonstrate that our framework not only achieves state-of-the-art accuracy across the DDD17-Seg, DSEC-Semantic, and M3ED-Semantic datasets but also significantly reduces energy consumption, achieving a 65\% reduction on the DSEC-Semantic dataset.

Peixi Wu, Bosong Chai, Menghua Zheng et al.

Bio-inspired Spiking Neural Networks (SNNs) provide an energy-efficient way to extract 3D spatio-temporal features. However, existing 3D SNNs have struggled with long-range dependencies until the recent emergence of Mamba, which offers superior computational efficiency and sequence modeling capability. In this work, we propose Spiking Point Mamba (SPM), the first Mamba-based SNN in the 3D domain. Due to the poor performance of simply transferring Mamba to 3D SNNs, SPM is designed to utilize both the sequence modeling capabilities of Mamba and the temporal feature extraction of SNNs. Specifically, we first introduce Hierarchical Dynamic Encoding (HDE), an improved direct encoding method that effectively introduces dynamic temporal mechanism, thereby facilitating temporal interactions. Then, we propose a Spiking Mamba Block (SMB), which builds upon Mamba while learning inter-time-step features and minimizing information loss caused by spikes. Finally, to further enhance model performance, we adopt an asymmetric SNN-ANN architecture for spike-based pre-training and finetune. Compared with the previous state-of-the-art SNN models, SPM improves OA by +6.2%, +6.1%, and +7.4% on three variants of ScanObjectNN, and boosts instance mIOU by +1.9% on ShapeNetPart. Meanwhile, its energy consumption is at least 3.5x lower than that of its ANN counterpart. The code will be made publicly available.

Yi Yang, Kefan Gu, Yuqing Wen et al.

While Vision-Language-Action (VLA) models have demonstrated impressive capabilities in robotic manipulation, their performance in complex reasoning and long-horizon task planning is limited by data scarcity and model capacity. To address this, we introduce ManiAgent, an agentic architecture for general manipulation tasks that achieves end-to-end output from task descriptions and environmental inputs to robotic manipulation actions. In this framework, multiple agents involve inter-agent communication to perform environmental perception, sub-task decomposition and action generation, enabling efficient handling of complex manipulation scenarios. Evaluations show ManiAgent achieves an 86.8% success rate on the SimplerEnv benchmark and 95.8% on real-world pick-and-place tasks, enabling efficient data collection that yields VLA models with performance comparable to those trained on human-annotated datasets. The project webpage is available at https://yi-yang929.github.io/ManiAgent/.

Wei Li, Hebei Li, Yansong Peng et al.

Diffusion models have significantly advanced text-to-image generation, laying the foundation for the development of personalized generative frameworks. However, existing methods lack precise layout controllability and overlook the potential of dynamic features of reference subjects in improving fidelity. In this work, we propose Layout-Controllable Personalized Diffusion (LCP-Diffusion) model, a novel framework that integrates subject identity preservation with flexible layout guidance in a tuning-free approach. Our model employs a Dynamic-Static Complementary Visual Refining module to comprehensively capture the intricate details of reference subjects, and introduces a Dual Layout Control mechanism to enforce robust spatial control across both training and inference stages. Extensive experiments validate that LCP-Diffusion excels in both identity preservation and layout controllability. To the best of our knowledge, this is a pioneering work enabling users to "create anything anywhere".

Peixi Wu, Bosong Chai, Xuan Nie et al.

In this technical report, we present our findings from the research conducted on the Vast Vocabulary Visual Detection (V3Det) dataset for Supervised Vast Vocabulary Visual Detection task. How to deal with complex categories and detection boxes has become a difficulty in this track. The original supervised detector is not suitable for this task. We have designed a series of improvements, including adjustments to the network structure, changes to the loss function, and design of training strategies. Our model has shown improvement over the baseline and achieved excellent rankings on the Leaderboard for both the Vast Vocabulary Object Detection (Supervised) track and the Open Vocabulary Object Detection (OVD) track of the V3Det Challenge 2024.