ResVG: Enhancing Relation and Semantic Understanding in Multiple Instances for Visual Grounding

Visual grounding aims to localize the object referred to in an image based on a natural language query. Although progress has been made recently, accurately localizing target objects within multiple-instance distractions (multiple objects of the same category as the target) remains a significant challenge. Existing methods demonstrate a significant performance drop when there are multiple distractions in an image, indicating an insufficient understanding of the fine-grained semantics and spatial relationships between objects. In this paper, we propose a novel approach, the Relation and Semantic-sensitive Visual Grounding (ResVG) model, to address this issue. Firstly, we enhance the model's understanding of fine-grained semantics by injecting semantic prior information derived from text queries into the model. This is achieved by leveraging text-to-image generation models to produce images representing the semantic attributes of target objects described in queries. Secondly, we tackle the lack of training samples with multiple distractions by introducing a relation-sensitive data augmentation method. This method generates additional training data by synthesizing images containing multiple objects of the same category and pseudo queries based on their spatial relationships. The proposed ReSVG model significantly improves the model's ability to comprehend both object semantics and spatial relations, leading to enhanced performance in visual grounding tasks, particularly in scenarios with multiple-instance distractions. We conduct extensive experiments to validate the effectiveness of our methods on five datasets. Code is available at https://github.com/minghangz/ResVG.

Training-free Video Temporal Grounding using Large-scale Pre-trained Models

Video temporal grounding aims to identify video segments within untrimmed videos that are most relevant to a given natural language query. Existing video temporal localization models rely on specific datasets for training and have high data collection costs, but they exhibit poor generalization capability under the across-dataset and out-of-distribution (OOD) settings. In this paper, we propose a Training-Free Video Temporal Grounding (TFVTG) approach that leverages the ability of pre-trained large models. A naive baseline is to enumerate proposals in the video and use the pre-trained visual language models (VLMs) to select the best proposal according to the vision-language alignment. However, most existing VLMs are trained on image-text pairs or trimmed video clip-text pairs, making it struggle to (1) grasp the relationship and distinguish the temporal boundaries of multiple events within the same video; (2) comprehend and be sensitive to the dynamic transition of events (the transition from one event to another) in the video. To address these issues, we propose leveraging large language models (LLMs) to analyze multiple sub-events contained in the query text and analyze the temporal order and relationships between these events. Secondly, we split a sub-event into dynamic transition and static status parts and propose the dynamic and static scoring functions using VLMs to better evaluate the relevance between the event and the description. Finally, for each sub-event description, we use VLMs to locate the top-k proposals and leverage the order and relationships between sub-events provided by LLMs to filter and integrate these proposals. Our method achieves the best performance on zero-shot video temporal grounding on Charades-STA and ActivityNet Captions datasets without any training and demonstrates better generalization capabilities in cross-dataset and OOD settings.

Poisson Ordinal Network for Gleason Group Estimation Using Bi-Parametric MRI

The Gleason groups serve as the primary histological grading system for prostate cancer, providing crucial insights into the cancer's potential for growth and metastasis. In clinical practice, pathologists determine the Gleason groups based on specimens obtained from ultrasound-guided biopsies. In this study, we investigate the feasibility of directly estimating the Gleason groups from MRI scans to reduce otherwise required biopsies. We identify two characteristics of this task, ordinality and the resulting dependent yet unknown variances between Gleason groups. In addition to the inter- / intra- observer variability in a multi-step Gleason scoring process based on the interpretation of Gleason patterns, our MR-based prediction is also subject to specimen sampling variance and, to a lesser degree, varying MR imaging protocols. To address this challenge, we propose a novel Poisson ordinal network (PON). PONs model the prediction using a Poisson distribution and leverages Poisson encoding and Poisson focal loss to capture a learnable dependency between ordinal classes (here, Gleason groups), rather than relying solely on the numerical ground-truth (e.g. Gleason Groups 1-5 or Gleason Scores 6-10). To improve this modelling efficacy, PONs also employ contrastive learning with a memory bank to regularise intra-class variance, decoupling the memory requirement of contrast learning from the batch size. Experimental results based on the images labelled by saturation biopsies from 265 prior-biopsy-blind patients, across two tasks demonstrate the superiority and effectiveness of our proposed method.

3D Vision and Language Pretraining with Large-Scale Synthetic Data

3D Vision-Language Pre-training (3D-VLP) aims to provide a pre-train model which can bridge 3D scenes with natural language, which is an important technique for embodied intelligence. However, current 3D-VLP datasets are hindered by limited scene-level diversity and insufficient fine-grained annotations (only 1.2K scenes and 280K textual annotations in ScanScribe), primarily due to the labor-intensive of collecting and annotating 3D scenes. To overcome these obstacles, we construct SynVL3D, a comprehensive synthetic scene-text corpus with 10K indoor scenes and 1M descriptions at object, view, and room levels, which has the advantages of diverse scene data, rich textual descriptions, multi-grained 3D-text associations, and low collection cost. Utilizing the rich annotations in SynVL3D, we pre-train a simple and unified Transformer for aligning 3D and language with multi-grained pretraining tasks. Moreover, we propose a synthetic-to-real domain adaptation in downstream task fine-tuning process to address the domain shift. Through extensive experiments, we verify the effectiveness of our model design by achieving state-of-the-art performance on downstream tasks including visual grounding, dense captioning, and question answering.

OED: Towards One-stage End-to-End Dynamic Scene Graph Generation

Dynamic Scene Graph Generation (DSGG) focuses on identifying visual relationships within the spatial-temporal domain of videos. Conventional approaches often employ multi-stage pipelines, which typically consist of object detection, temporal association, and multi-relation classification. However, these methods exhibit inherent limitations due to the separation of multiple stages, and independent optimization of these sub-problems may yield sub-optimal solutions. To remedy these limitations, we propose a one-stage end-to-end framework, termed OED, which streamlines the DSGG pipeline. This framework reformulates the task as a set prediction problem and leverages pair-wise features to represent each subject-object pair within the scene graph. Moreover, another challenge of DSGG is capturing temporal dependencies, we introduce a Progressively Refined Module (PRM) for aggregating temporal context without the constraints of additional trackers or handcrafted trajectories, enabling end-to-end optimization of the network. Extensive experiments conducted on the Action Genome benchmark demonstrate the effectiveness of our design. The code and models are available at \url{https://github.com/guanw-pku/OED}.

MERIT: Multi-view Evidential learning for Reliable and Interpretable liver fibrosis sTaging

Accurate staging of liver fibrosis from magnetic resonance imaging (MRI) is crucial in clinical practice. While conventional methods often focus on a specific sub-region, multi-view learning captures more information by analyzing multiple patches simultaneously. However, previous multi-view approaches could not typically calculate uncertainty by nature, and they generally integrate features from different views in a black-box fashion, hence compromising reliability as well as interpretability of the resulting models. In this work, we propose a new multi-view method based on evidential learning, referred to as MERIT, which tackles the two challenges in a unified framework. MERIT enables uncertainty quantification of the predictions to enhance reliability, and employs a logic-based combination rule to improve interpretability. Specifically, MERIT models the prediction from each sub-view as an opinion with quantified uncertainty under the guidance of the subjective logic theory. Furthermore, a distribution-aware base rate is introduced to enhance performance, particularly in scenarios involving class distribution shifts. Finally, MERIT adopts a feature-specific combination rule to explicitly fuse multi-view predictions, thereby enhancing interpretability. Results have showcased the effectiveness of the proposed MERIT, highlighting the reliability and offering both ad-hoc and post-hoc interpretability. They also illustrate that MERIT can elucidate the significance of each view in the decision-making process for liver fibrosis staging.

Progressive trajectory matching for medical dataset distillation

It is essential but challenging to share medical image datasets due to privacy issues, which prohibit building foundation models and knowledge transfer. In this paper, we propose a novel dataset distillation method to condense the original medical image datasets into a synthetic one that preserves useful information for building an analysis model without accessing the original datasets. Existing methods tackle only natural images by randomly matching parts of the training trajectories of the model parameters trained by the whole real datasets. However, through extensive experiments on medical image datasets, the training process is extremely unstable and achieves inferior distillation results. To solve these barriers, we propose to design a novel progressive trajectory matching strategy to improve the training stability for medical image dataset distillation. Additionally, it is observed that improved stability prevents the synthetic dataset diversity and final performance improvements. Therefore, we propose a dynamic overlap mitigation module that improves the synthetic dataset diversity by dynamically eliminating the overlap across different images and retraining parts of the synthetic images for better convergence. Finally, we propose a new medical image dataset distillation benchmark of various modalities and configurations to promote fair evaluations. It is validated that our proposed method achieves 8.33% improvement over previous state-of-the-art methods on average, and 11.7% improvement when ipc=2 (i.e., image per class is 2). Codes and benchmarks will be released.

On the Model-Agnostic Multi-Source-Free Unsupervised Domain Adaptation

Multi-Source-Free Unsupervised Domain Adaptation (MSFDA) aims to transfer knowledge from multiple well-labeled source domains to an unlabeled target domain, using source models instead of source data. Existing MSFDA methods limited that each source domain provides only a single model, with a uniform structure. This paper introduces a new MSFDA setting: Model-Agnostic Multi-Source-Free Unsupervised Domain Adaptation (MMDA), allowing diverse source models with varying architectures, without quantitative restrictions. While MMDA holds promising potential, incorporating numerous source models poses a high risk of including undesired models, which highlights the source model selection problem. To address it, we first provide a theoretical analysis of this problem. We reveal two fundamental selection principles: transferability principle and diversity principle, and introduce a selection algorithm to integrate them. Then, considering the measure of transferability is challenging, we propose a novel Source-Free Unsupervised Transferability Estimation (SUTE). This novel formulation enables the assessment and comparison of transferability across multiple source models with different architectures in the context of domain shift, without requiring access to any target labels or source data. Based on the above, we introduce a new framework to address MMDA. Specifically, we first conduct source model selection based on the proposed selection principles. Subsequently, we design two modules to aggregate knowledge from included models and recycle useful knowledge from excluded models. These modules enable us to leverage source knowledge efficiently and effectively, thereby supporting us in learning a discriminative target model via adaptation. We validate the effectiveness of our method through numerous experimental results, and demonstrate that our approach achieves state-of-the-art performance.

EviPrompt: A Training-Free Evidential Prompt Generation Method for Segment Anything Model in Medical Images

Medical image segmentation has immense clinical applicability but remains a challenge despite advancements in deep learning. The Segment Anything Model (SAM) exhibits potential in this field, yet the requirement for expertise intervention and the domain gap between natural and medical images poses significant obstacles. This paper introduces a novel training-free evidential prompt generation method named EviPrompt to overcome these issues. The proposed method, built on the inherent similarities within medical images, requires only a single reference image-annotation pair, making it a training-free solution that significantly reduces the need for extensive labeling and computational resources. First, to automatically generate prompts for SAM in medical images, we introduce an evidential method based on uncertainty estimation without the interaction of clinical experts. Then, we incorporate the human prior into the prompts, which is vital for alleviating the domain gap between natural and medical images and enhancing the applicability and usefulness of SAM in medical scenarios. EviPrompt represents an efficient and robust approach to medical image segmentation, with evaluations across a broad range of tasks and modalities confirming its efficacy.

Efficient Adaptive Human-Object Interaction Detection with Concept-guided Memory

Human Object Interaction (HOI) detection aims to localize and infer the relationships between a human and an object. Arguably, training supervised models for this task from scratch presents challenges due to the performance drop over rare classes and the high computational cost and time required to handle long-tailed distributions of HOIs in complex HOI scenes in realistic settings. This observation motivates us to design an HOI detector that can be trained even with long-tailed labeled data and can leverage existing knowledge from pre-trained models. Inspired by the powerful generalization ability of the large Vision-Language Models (VLM) on classification and retrieval tasks, we propose an efficient Adaptive HOI Detector with Concept-guided Memory (ADA-CM). ADA-CM has two operating modes. The first mode makes it tunable without learning new parameters in a training-free paradigm. Its second mode incorporates an instance-aware adapter mechanism that can further efficiently boost performance if updating a lightweight set of parameters can be afforded. Our proposed method achieves competitive results with state-of-the-art on the HICO-DET and V-COCO datasets with much less training time. Code can be found at https://github.com/ltttpku/ADA-CM.