arXiv subjects

Xiangyang Xue

Publications and source records attributed to Xiangyang Xue.

At least 73 records · Page 4Linked to original sources

arXiv2025

Towards Enhanced Image Inpainting: Mitigating Unwanted Object Insertion and Preserving Color Consistency

Recent advances in image inpainting increasingly use generative models to handle large irregular masks. However, these models can create unrealistic inpainted images due to two main issues: (1) Unwanted object insertion: Even with unmasked areas as context, generative models may still generate arbitrary objects in the masked region that don't align with the rest of the image. (2) Color inconsistency: Inpainted regions often have color shifts that causes a smeared appearance, reducing image quality. Retraining the generative model could help solve these issues, but it's costly since state-of-the-art latent-based diffusion and rectified flow models require a three-stage training process: training a VAE, training a generative U-Net or transformer, and fine-tuning for inpainting. Instead, this paper proposes a post-processing approach, dubbed as ASUKA (Aligned Stable inpainting with UnKnown Areas prior), to improve inpainting models. To address unwanted object insertion, we leverage a Masked Auto-Encoder (MAE) for reconstruction-based priors. This mitigates object hallucination while maintaining the model's generation capabilities. To address color inconsistency, we propose a specialized VAE decoder that treats latent-to-image decoding as a local harmonization task, significantly reducing color shifts for color-consistent inpainting. We validate ASUKA on SD 1.5 and FLUX inpainting variants with Places2 and MISATO, our proposed diverse collection of datasets. Results show that ASUKA mitigates object hallucination and improves color consistency over standard diffusion and rectified flow models and other inpainting methods.

cs.CV↗

arXiv2025

Towards a Vision-Language Episodic Memory Framework: Large-scale Pretrained Model-Augmented Hippocampal Attractor Dynamics

Modeling episodic memory (EM) remains a significant challenge in both neuroscience and AI, with existing models either lacking interpretability or struggling with practical applications. This paper proposes the Vision-Language Episodic Memory (VLEM) framework to address these challenges by integrating large-scale pretrained models with hippocampal attractor dynamics. VLEM leverages the strong semantic understanding of pretrained models to transform sensory input into semantic embeddings as the neocortex, while the hippocampus supports stable memory storage and retrieval through attractor dynamics. In addition, VLEM incorporates prefrontal working memory and the entorhinal gateway, allowing interaction between the neocortex and the hippocampus. To facilitate real-world applications, we introduce EpiGibson, a 3D simulation platform for generating episodic memory data. Experimental results demonstrate the VLEM framework's ability to efficiently learn high-level temporal representations from sensory input, showcasing its robustness, interpretability, and applicability in real-world scenarios.

q-bio.NC↗

arXiv2025

One-Shot Heterogeneous Federated Learning with Local Model-Guided Diffusion Models

In recent years, One-shot Federated Learning methods based on Diffusion Models have garnered increasing attention due to their remarkable performance. However, most of these methods require the deployment of foundation models on client devices, which significantly raises the computational requirements and reduces their adaptability to heterogeneous client models compared to traditional FL methods. In this paper, we propose FedLMG, a heterogeneous one-shot Federated learning method with Local Model-Guided diffusion models. Briefly speaking, in FedLMG, clients do not need access to any foundation models but only train and upload their local models, which is consistent with traditional FL methods. On the clients, we employ classification loss and BN loss to capture the broad category features and detailed contextual features of the client distributions. On the server, based on the uploaded client models, we utilize backpropagation to guide the server's DM in generating synthetic datasets that comply with the client distributions, which are then used to train the aggregated model. By using the locally trained client models as a medium to transfer client knowledge, our method significantly reduces the computational requirements on client devices and effectively adapts to scenarios with heterogeneous clients. Extensive quantitation and visualization experiments on three large-scale real-world datasets, along with theoretical analysis, demonstrate that the synthetic datasets generated by FedLMG exhibit comparable quality and diversity to the client datasets, which leads to an aggregated model that outperforms all compared methods and even the performance ceiling, further elucidating the significant potential of utilizing DMs in FL.

cs.CV↗

arXiv2025

EmoHead: Emotional Talking Head via Manipulating Semantic Expression Parameters

Generating emotion-specific talking head videos from audio input is an important and complex challenge for human-machine interaction. However, emotion is highly abstract concept with ambiguous boundaries, and it necessitates disentangled expression parameters to generate emotionally expressive talking head videos. In this work, we present EmoHead to synthesize talking head videos via semantic expression parameters. To predict expression parameter for arbitrary audio input, we apply an audio-expression module that can be specified by an emotion tag. This module aims to enhance correlation from audio input across various emotions. Furthermore, we leverage pre-trained hyperplane to refine facial movements by probing along the vertical direction. Finally, the refined expression parameters regularize neural radiance fields and facilitate the emotion-consistent generation of talking head videos. Experimental results demonstrate that semantic expression parameters lead to better reconstruction quality and controllability.

cs.CV↗

arXiv2025

DecoFuse: Decomposing and Fusing the "What", "Where", and "How" for Brain-Inspired fMRI-to-Video Decoding

Decoding visual experiences from brain activity is a significant challenge. Existing fMRI-to-video methods often focus on semantic content while overlooking spatial and motion information. However, these aspects are all essential and are processed through distinct pathways in the brain. Motivated by this, we propose DecoFuse, a novel brain-inspired framework for decoding videos from fMRI signals. It first decomposes the video into three components - semantic, spatial, and motion - then decodes each component separately before fusing them to reconstruct the video. This approach not only simplifies the complex task of video decoding by decomposing it into manageable sub-tasks, but also establishes a clearer connection between learned representations and their biological counterpart, as supported by ablation studies. Further, our experiments show significant improvements over previous state-of-the-art methods, achieving 82.4% accuracy for semantic classification, 70.6% accuracy in spatial consistency, a 0.212 cosine similarity for motion prediction, and 21.9% 50-way accuracy for video generation. Additionally, neural encoding analyses for semantic and spatial information align with the two-streams hypothesis, further validating the distinct roles of the ventral and dorsal pathways. Overall, DecoFuse provides a strong and biologically plausible framework for fMRI-to-video decoding. Project page: https://chongjg.github.io/DecoFuse/.

cs.CV↗

arXiv2025

ReasonGrounder: LVLM-Guided Hierarchical Feature Splatting for Open-Vocabulary 3D Visual Grounding and Reasoning

Open-vocabulary 3D visual grounding and reasoning aim to localize objects in a scene based on implicit language descriptions, even when they are occluded. This ability is crucial for tasks such as vision-language navigation and autonomous robotics. However, current methods struggle because they rely heavily on fine-tuning with 3D annotations and mask proposals, which limits their ability to handle diverse semantics and common knowledge required for effective reasoning. In this work, we propose ReasonGrounder, an LVLM-guided framework that uses hierarchical 3D feature Gaussian fields for adaptive grouping based on physical scale, enabling open-vocabulary 3D grounding and reasoning. ReasonGrounder interprets implicit instructions using large vision-language models (LVLM) and localizes occluded objects through 3D Gaussian splatting. By incorporating 2D segmentation masks from the SAM and multi-view CLIP embeddings, ReasonGrounder selects Gaussian groups based on object scale, enabling accurate localization through both explicit and implicit language understanding, even in novel, occluded views. We also contribute ReasoningGD, a new dataset containing over 10K scenes and 2 million annotations for evaluating open-vocabulary 3D grounding and amodal perception under occlusion. Experiments show that ReasonGrounder significantly improves 3D grounding accuracy in real-world scenarios.

cs.CV↗

arXiv2025

Multi-modality Anomaly Segmentation on the Road

Semantic segmentation allows autonomous driving cars to understand the surroundings of the vehicle comprehensively. However, it is also crucial for the model to detect obstacles that may jeopardize the safety of autonomous driving systems. Based on our experiments, we find that current uni-modal anomaly segmentation frameworks tend to produce high anomaly scores for non-anomalous regions in images. Motivated by this empirical finding, we develop a multi-modal uncertainty-based anomaly segmentation framework, named MMRAS+, for autonomous driving systems. MMRAS+ effectively reduces the high anomaly outputs of non-anomalous classes by introducing text-modal using the CLIP text encoder. Indeed, MMRAS+ is the first multi-modal anomaly segmentation solution for autonomous driving. Moreover, we develop an ensemble module to further boost the anomaly segmentation performance. Experiments on RoadAnomaly, SMIYC, and Fishyscapes validation datasets demonstrate the superior performance of our method. The code is available in https://github.com/HengGao12/MMRAS_plus.

cs.CV↗

arXiv2025

Content and Salient Semantics Collaboration for Cloth-Changing Person Re-Identification

Cloth-changing person re-identification aims at recognizing the same person with clothing changes across non-overlapping cameras. Advanced methods either resort to identity-related auxiliary modalities (e.g., sketches, silhouettes, and keypoints) or clothing labels to mitigate the impact of clothes. However, relying on unpractical and inflexible auxiliary modalities or annotations limits their real-world applicability. In this paper, we promote cloth-changing person re-identification by leveraging abundant semantics present within pedestrian images, without the need for any auxiliaries. Specifically, we first propose a unified Semantics Mining and Refinement (SMR) module to extract robust identity-related content and salient semantics, mitigating interference from clothing appearances effectively. We further propose the Content and Salient Semantics Collaboration (CSSC) framework to collaborate and leverage various semantics, facilitating cross-parallel semantic interaction and refinement. Our proposed method achieves state-of-the-art performance on three cloth-changing benchmarks, demonstrating its superiority over advanced competitors. The code is available at https://github.com/QizaoWang/CSSC-CCReID.

cs.CV↗

arXiv2025

MVGenMaster: Scaling Multi-View Generation from Any Image via 3D Priors Enhanced Diffusion Model

We introduce MVGenMaster, a multi-view diffusion model enhanced with 3D priors to address versatile Novel View Synthesis (NVS) tasks. MVGenMaster leverages 3D priors that are warped using metric depth and camera poses, significantly enhancing both generalization and 3D consistency in NVS. Our model features a simple yet effective pipeline that can generate up to 100 novel views conditioned on variable reference views and camera poses with a single forward process. Additionally, we have developed a comprehensive large-scale multi-view image dataset called MvD-1M, comprising up to 1.6 million scenes, equipped with well-aligned metric depth to train MVGenMaster. Moreover, we present several training and model modifications to strengthen the model with scaled-up datasets. Extensive evaluations across in- and out-of-domain benchmarks demonstrate the effectiveness of our proposed method and data formulation. Models and codes will be released at https://github.com/ewrfcas/MVGenMaster/.

cs.CV↗

arXiv2025

AnyRefill: A Unified, Data-Efficient Framework for Left-Prompt-Guided Vision Tasks

In this paper, we present a novel Left-Prompt-Guided (LPG) paradigm to address a diverse range of reference-based vision tasks. Inspired by the human creative process, we reformulate these tasks using a left-right stitching formulation to construct contextual input. Building upon this foundation, we propose AnyRefill, an extension of LeftRefill, that effectively adapts Text-to-Image (T2I) models to various vision tasks. AnyRefill leverages the inpainting priors of advanced T2I model based on the Diffusion Transformer (DiT) architecture, and incorporates flexible components to enhance its capabilities. By combining task-specific LoRAs with the stitching input, AnyRefill unlocks its potential across diverse tasks, including conditional generation, visual perception, and image editing, without requiring additional visual encoders. Meanwhile, AnyRefill exhibits remarkable data efficiency, requiring minimal task-specific fine-tuning while maintaining high generative performance. Through extensive ablation studies, we demonstrate that AnyRefill outperforms other image condition injection methods and achieves competitive results compared to state-of-the-art open-source methods. Notably, AnyRefill delivers results comparable to advanced commercial tools, such as IC-Light and SeedEdit, even in challenging scenarios. Comprehensive experiments and ablation studies across versatile tasks validate the strong generation of the proposed simple yet effective LPG formulation, establishing AnyRefill as a unified, highly data-efficient solution for reference-based vision tasks.

cs.CV↗

arXiv2025

Distribution Aligned Semantics Adaption for Lifelong Person Re-Identification

In real-world scenarios, person Re-IDentification (Re-ID) systems need to be adaptable to changes in space and time. Therefore, the adaptation of Re-ID models to new domains while preserving previously acquired knowledge is crucial, known as Lifelong person Re-IDentification (LReID). Advanced LReID methods rely on replaying exemplars from old domains and applying knowledge distillation in logits with old models. However, due to privacy concerns, retaining previous data is inappropriate. Additionally, the fine-grained and open-set characteristics of Re-ID limit the effectiveness of the distillation paradigm for accumulating knowledge. We argue that a Re-ID model trained on diverse and challenging pedestrian images at a large scale can acquire robust and general human semantic knowledge. These semantics can be readily utilized as shared knowledge for lifelong applications. In this paper, we identify the challenges and discrepancies associated with adapting a pre-trained model to each application domain and introduce the Distribution Aligned Semantics Adaption (DASA) framework. It efficiently adjusts Batch Normalization (BN) to mitigate interference from data distribution discrepancy and freezes the pre-trained convolutional layers to preserve shared knowledge. Additionally, we propose the lightweight Semantics Adaption (SA) module, which effectively adapts learned semantics to enhance pedestrian representations. Extensive experiments demonstrate the remarkable superiority of our proposed framework over advanced LReID methods, and it exhibits significantly reduced storage consumption. DASA presents a novel and cost-effective perspective on effectively adapting pre-trained models for LReID. The code is available at https://github.com/QizaoWang/DASA-LReID.

cs.CV↗

arXiv2025

Global Semantic-Guided Sub-image Feature Weight Allocation in High-Resolution Large Vision-Language Models

As the demand for high-resolution image processing in Large Vision-Language Models (LVLMs) grows, sub-image partitioning has become a popular approach for mitigating visual information loss associated with fixed-resolution processing. However, existing partitioning methods uniformly process sub-images, resulting in suboptimal image understanding. In this work, we reveal that the sub-images with higher semantic relevance to the entire image encapsulate richer visual information for preserving the model's visual understanding ability. Therefore, we propose the Global Semantic-guided Weight Allocator (GSWA) module, which dynamically allocates weights to sub-images based on their relative information density, emulating human visual attention mechanisms. This approach enables the model to focus on more informative regions, overcoming the limitations of uniform treatment. We integrate GSWA into the InternVL2-2B framework to create SleighVL, a lightweight yet high-performing model. Extensive experiments demonstrate that SleighVL outperforms models with comparable parameters and remains competitive with larger models. Our work provides a promising direction for more efficient and contextually aware high-resolution image processing in LVLMs, advancing multimodal system development.

cs.CV↗

arXiv2025

Instruction-Guided Fusion of Multi-Layer Visual Features in Large Vision-Language Models

Large Vision-Language Models (LVLMs) have achieved remarkable success in a wide range of multimodal tasks by integrating pre-trained vision encoders and large language models. However, current LVLMs primarily rely on visual features extracted from the final layers of the vision encoder, overlooking the complementary information available in shallower layers. While recent approaches have explored the use of multilayer visual features in LVLMs, they tend to be task-agnostic and fail to examine the dependencies of hierarchical visual features on specific tasks. To address these gaps, we systematically investigate the contributions of visual features from different encoder layers using 18 benchmarks spanning 6 task categories. Our findings reveal that multilayer features provide complementary strengths with varying task dependencies, and uniform fusion leads to suboptimal performance. Building on these insights, we propose the instruction-guided vision aggregator, a module that dynamically integrates multi-layer visual features based on textual instructions, without increasing the number of visual tokens. Extensive evaluations demonstrate the superior performance of our method. Additionally, an in-depth analysis of the aggregator's behavior highlights the dominance of mid-to-high-level features in semantic-rich tasks and the critical role of low-level features in fine-grained perception.

cs.CV↗

arXiv2024

Topo-Field: Topometric mapping with Brain-inspired Hierarchical Layout-Object-Position Fields

Mobile robots require comprehensive scene understanding to operate effectively in diverse environments, enriched with contextual information such as layouts, objects, and their relationships. Although advances like neural radiation fields (NeRFs) offer high-fidelity 3D reconstructions, they are computationally intensive and often lack efficient representations of traversable spaces essential for planning and navigation. In contrast, topological maps are computationally efficient but lack the semantic richness necessary for a more complete understanding of the environment. Inspired by a population code in the postrhinal cortex (POR) strongly tuned to spatial layouts over scene content rapidly forming a high-level cognitive map, this work introduces Topo-Field, a framework that integrates Layout-Object-Position (LOP) associations into a neural field and constructs a topometric map from this learned representation. LOP associations are modeled by explicitly encoding object and layout information, while a Large Foundation Model (LFM) technique allows for efficient training without extensive annotations. The topometric map is then constructed by querying the learned neural representation, offering both semantic richness and computational efficiency. Empirical evaluations in multi-room environments demonstrate the effectiveness of Topo-Field in tasks such as position attribute inference, query localization, and topometric planning, successfully bridging the gap between high-fidelity scene understanding and efficient robotic navigation.

cs.RO↗

arXiv2024

Automated Label Unification for Multi-Dataset Semantic Segmentation with GNNs

Deep supervised models possess significant capability to assimilate extensive training data, thereby presenting an opportunity to enhance model performance through training on multiple datasets. However, conflicts arising from different label spaces among datasets may adversely affect model performance. In this paper, we propose a novel approach to automatically construct a unified label space across multiple datasets using graph neural networks. This enables semantic segmentation models to be trained simultaneously on multiple datasets, resulting in performance improvements. Unlike existing methods, our approach facilitates seamless training without the need for additional manual reannotation or taxonomy reconciliation. This significantly enhances the efficiency and effectiveness of multi-dataset segmentation model training. The results demonstrate that our method significantly outperforms other multi-dataset training methods when trained on seven datasets simultaneously, and achieves state-of-the-art performance on the WildDash 2 benchmark.

cs.CV↗

arXiv2024

SparseGrasp: Robotic Grasping via 3D Semantic Gaussian Splatting from Sparse Multi-View RGB Images

Language-guided robotic grasping is a rapidly advancing field where robots are instructed using human language to grasp specific objects. However, existing methods often depend on dense camera views and struggle to quickly update scenes, limiting their effectiveness in changeable environments. In contrast, we propose SparseGrasp, a novel open-vocabulary robotic grasping system that operates efficiently with sparse-view RGB images and handles scene updates fastly. Our system builds upon and significantly enhances existing computer vision modules in robotic learning. Specifically, SparseGrasp utilizes DUSt3R to generate a dense point cloud as the initialization for 3D Gaussian Splatting (3DGS), maintaining high fidelity even under sparse supervision. Importantly, SparseGrasp incorporates semantic awareness from recent vision foundation models. To further improve processing efficiency, we repurpose Principal Component Analysis (PCA) to compress features from 2D models. Additionally, we introduce a novel render-and-compare strategy that ensures rapid scene updates, enabling multi-turn grasping in changeable environments. Experimental results show that SparseGrasp significantly outperforms state-of-the-art methods in terms of both speed and adaptability, providing a robust solution for multi-turn grasping in changeable environment.

cs.RO↗

arXiv2024

Repositioning the Subject within Image

Current image manipulation primarily centers on static manipulation, such as replacing specific regions within an image or altering its overall style. In this paper, we introduce an innovative dynamic manipulation task, subject repositioning. This task involves relocating a user-specified subject to a desired position while preserving the image's fidelity. Our research reveals that the fundamental sub-tasks of subject repositioning, which include filling the void left by the repositioned subject, reconstructing obscured portions of the subject and blending the subject to be consistent with surrounding areas, can be effectively reformulated as a unified, prompt-guided inpainting task. Consequently, we can employ a single diffusion generative model to address these sub-tasks using various task prompts learned through our proposed task inversion technique. Additionally, we integrate pre-processing and post-processing techniques to further enhance the quality of subject repositioning. These elements together form our SEgment-gEnerate-and-bLEnd (SEELE) framework. To assess SEELE's effectiveness in subject repositioning, we assemble a real-world subject repositioning dataset called ReS. Results of SEELE on ReS demonstrate its efficacy. Code and ReS dataset are available at https://yikai-wang.github.io/seele/.

cs.CV↗

arXiv2024

MVInpainter: Learning Multi-View Consistent Inpainting to Bridge 2D and 3D Editing

Novel View Synthesis (NVS) and 3D generation have recently achieved prominent improvements. However, these works mainly focus on confined categories or synthetic 3D assets, which are discouraged from generalizing to challenging in-the-wild scenes and fail to be employed with 2D synthesis directly. Moreover, these methods heavily depended on camera poses, limiting their real-world applications. To overcome these issues, we propose MVInpainter, re-formulating the 3D editing as a multi-view 2D inpainting task. Specifically, MVInpainter partially inpaints multi-view images with the reference guidance rather than intractably generating an entirely novel view from scratch, which largely simplifies the difficulty of in-the-wild NVS and leverages unmasked clues instead of explicit pose conditions. To ensure cross-view consistency, MVInpainter is enhanced by video priors from motion components and appearance guidance from concatenated reference key&value attention. Furthermore, MVInpainter incorporates slot attention to aggregate high-level optical flow features from unmasked regions to control the camera movement with pose-free training and inference. Sufficient scene-level experiments on both object-centric and forward-facing datasets verify the effectiveness of MVInpainter, including diverse tasks, such as multi-view object removal, synthesis, insertion, and replacement. The project page is https://ewrfcas.github.io/MVInpainter/.

cs.CV↗