Kewei Wang

Chengxin Liu, Kewei Wang, Hao Lu et al.

Learning accurate object detectors often requires large-scale training data with precise object bounding boxes. However, labeling such data is expensive and time-consuming. As the crowd-sourcing labeling process and the ambiguities of the objects may raise noisy bounding box annotations, the object detectors will suffer from the degenerated training data. In this work, we aim to address the challenge of learning robust object detectors with inaccurate bounding boxes. Inspired by the fact that localization precision suffers significantly from inaccurate bounding boxes while classification accuracy is less affected, we propose leveraging classification as a guidance signal for refining localization results. Specifically, by treating an object as a bag of instances, we introduce an Object-Aware Multiple Instance Learning approach (OA-MIL), featured with object-aware instance selection and object-aware instance extension. The former aims to select accurate instances for training, instead of directly using inaccurate box annotations. The latter focuses on generating high-quality instances for selection. Extensive experiments on synthetic noisy datasets (i.e., noisy PASCAL VOC and MS-COCO) and a real noisy wheat head dataset demonstrate the effectiveness of our OA-MIL. Code is available at https://github.com/cxliu0/OA-MIL.

Jiwei Chen, Kewei Wang, Wen Su et al.

Crowd counting based on density maps is generally regarded as a regression task.Deep learning is used to learn the mapping between image content and crowd density distribution. Although great success has been achieved, some pedestrians far away from the camera are difficult to be detected. And the number of hard examples is often larger. Existing methods with simple Euclidean distance algorithm indiscriminately optimize the hard and easy examples so that the densities of hard examples are usually incorrectly predicted to be lower or even zero, which results in large counting errors. To address this problem, we are the first to propose the Hard Example Focusing(HEF) algorithm for the regression task of crowd counting. The HEF algorithm makes our model rapidly focus on hard examples by attenuating the contribution of easy examples.Then higher importance will be given to the hard examples with wrong estimations. Moreover, the scale variations in crowd scenes are large, and the scale annotations are labor-intensive and expensive. By proposing a multi-Scale Semantic Refining (SSR) strategy, lower layers of our model can break through the limitation of deep learning to capture semantic features of different scales to sufficiently deal with the scale variation. We perform extensive experiments on six benchmark datasets to verify the proposed method. Results indicate the superiority of our proposed method over the state-of-the-art methods. Moreover, our designed model is smaller and faster.

Xiang Feng, Jiawei Zhou, Zhangfeng Huang et al.

Evaluating whether Multimodal Large Language Models can produce trustworthy, verifiable reasoning over long, visually rich documents requires evaluation beyond end-to-end answer accuracy. We introduce DocScope, a benchmark that formulates long-document QA as a structured reasoning trajectory prediction problem: given a complete PDF document and a question, the model outputs evidence pages, supporting evidence regions, relevant factual statements, and a final answer. We design a four-stage evaluation protocol -- Page Localization, Region Grounding, Fact Extraction, and Answer Verification -- that audits each level of the trajectory independently through inter-stage decoupling, with all judges selected and calibrated via human alignment studies. DocScope comprises 1,124 questions derived from 273 documents, with all hierarchical evidence annotations completed by human annotators. We benchmark 6 proprietary models, 12 open-weight models, and several domain-specific systems. Our experiments reveal that answer accuracy cannot substitute for trajectory-level evaluation: even among correct answers, the highest observed rate of complete evidence chains is only 29\%. Across all models, region grounding remains the weakest trajectory stage. Furthermore, the primary difficulty stems from aggregating evidence dispersed across long distances and multiple document clusters, while an oracle study identifies faithful perception and fact extraction as the dominant capability bottleneck. Cross-architecture comparisons further suggest that activated parameter count matters more than total scale. The benchmark and code will be publicly released at https://github.com/MiliLab/DocScope.

Kewei Wang, Yizheng Wu, Zhiyu Pan et al.

Class-agnostic motion prediction methods aim to comprehend motion within open-world scenarios, holding significance for autonomous driving systems. However, training a high-performance model in a fully-supervised manner always requires substantial amounts of manually annotated data, which can be both expensive and time-consuming to obtain. To address this challenge, our study explores the potential of semi-supervised learning (SSL) for class-agnostic motion prediction. Our SSL framework adopts a consistency-based self-training paradigm, enabling the model to learn from unlabeled data by generating pseudo labels through test-time inference. To improve the quality of pseudo labels, we propose a novel motion selection and re-generation module. This module effectively selects reliable pseudo labels and re-generates unreliable ones. Furthermore, we propose two data augmentation strategies: temporal sampling and BEVMix. These strategies facilitate consistency regularization in SSL. Experiments conducted on nuScenes demonstrate that our SSL method can surpass the self-supervised approach by a large margin by utilizing only a tiny fraction of labeled data. Furthermore, our method exhibits comparable performance to weakly and some fully supervised methods. These results highlight the ability of our method to strike a favorable balance between annotation costs and performance. Code will be available at https://github.com/kwwcv/SSMP.

Jun Cen, Yizheng Wu, Kewei Wang et al.

The Segment Anything Model (SAM) has demonstrated its effectiveness in segmenting any part of 2D RGB images. However, SAM exhibits a stronger emphasis on texture information while paying less attention to geometry information when segmenting RGB images. To address this limitation, we propose the Segment Any RGBD (SAD) model, which is specifically designed to extract geometry information directly from images. Inspired by the natural ability of humans to identify objects through the visualization of depth maps, SAD utilizes SAM to segment the rendered depth map, thus providing cues with enhanced geometry information and mitigating the issue of over-segmentation. We further include the open-vocabulary semantic segmentation in our framework, so that the 3D panoptic segmentation is fulfilled. The project is available on https://github.com/Jun-CEN/SegmentAnyRGBD.

Xingyi Li, Zhiguo Cao, Yizheng Wu et al.

Current 3D stylization methods often assume static scenes, which violates the dynamic nature of our real world. To address this limitation, we present S-DyRF, a reference-based spatio-temporal stylization method for dynamic neural radiance fields. However, stylizing dynamic 3D scenes is inherently challenging due to the limited availability of stylized reference images along the temporal axis. Our key insight lies in introducing additional temporal cues besides the provided reference. To this end, we generate temporal pseudo-references from the given stylized reference. These pseudo-references facilitate the propagation of style information from the reference to the entire dynamic 3D scene. For coarse style transfer, we enforce novel views and times to mimic the style details present in pseudo-references at the feature level. To preserve high-frequency details, we create a collection of stylized temporal pseudo-rays from temporal pseudo-references. These pseudo-rays serve as detailed and explicit stylization guidance for achieving fine style transfer. Experiments on both synthetic and real-world datasets demonstrate that our method yields plausible stylized results of space-time view synthesis on dynamic 3D scenes.