检索结果-内蒙古大学图书馆

Simple Image-level Classification Improves Open-vocabulary Object Detection

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Fang, Ruohuan Pang, Guansong Bai, Xiao School of Computer Science and Engineering Beihang University China School of Computing and Information Systems Singapore Management University Singapore State Key Laboratory of Software Development Environment Jiangxi Research Institute Beihang University China

Open-Vocabulary Object Detection (OVOD) aims to detect novel objects beyond a given set of base categories on which the detection model is trained. Recent OVOD methods focus on adapting the image-level pre-trained vision-language models (VLMs), such as CLIP, to a region-level object detection task via, e.g., region-level knowledge distillation, regional prompt learning, or region-text pre-training, to expand the detection vocabulary. These methods have demonstrated remarkable performance in recognizing regional visual concepts, but they are weak in exploiting the VLMs' powerful global scene understanding ability learned from the billion-scale image-level text descriptions. This limits their capability in detecting hard objects of small, blurred, or occluded appearance from novel/base categories, whose detection heavily relies on contextual information. To address this, we propose a novel approach, namely Simple Image-level Classification for Context-Aware Detection Scoring (SIC-CADS), to leverage the superior global knowledge yielded from CLIP for complementing the current OVOD models from a global perspective. The core of SIC-CADS is a multi-modal multi-label recognition (MLR) module that learns the object co-occurrence-based contextual information from CLIP to recognize all possible object categories in the scene. These image-level MLR scores can then be utilized to refine the instance-level detection scores of the current OVOD models in detecting those hard objects. This is verified by extensive empirical results on two popular benchmarks, OV-LVIS and OV-COCO, which show that SIC-CADS achieves significant and consistent improvement when combined with different types of OVOD models. Further, SIC-CADS also improves the cross-dataset generalization ability on Objects365 and OpenImages. Code is available at https://***/mala-lab/SIC-CADS. Copyright © 2023, The Authors. All rights reserved.

关键词： Object detection

Out-of-Distribution Detection in Long-Tailed Recognition with Calibrated Outlier Class Learning

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Miao, Wenjun Pang, Guansong Li, Tianqi Bai, Xiao Zheng, Jin School of Computer Science and Engineering Beihang University China School of Computing and Information Systems Singapore Management University Singapore State Key Laboratory of Software Development Environment Jiangxi Research Institute Beihang University China State Key Laboratory of Virtual Reality Technology and Systems Beihang University China

Existing out-of-distribution (OOD) methods have shown great success on balanced datasets but become ineffective in long-tailed recognition (LTR) scenarios where 1) OOD samples are often wrongly classified into head classes and/or 2) tail-class samples are treated as OOD samples. To address these issues, current studies fit a prior distribution of auxiliary/pseudo OOD data to the long-tailed in-distribution (ID) data. However, it is difficult to obtain such an accurate prior distribution given the unknowingness of real OOD samples and heavy class imbalance in LTR. A straightforward solution to avoid the requirement of this prior is to learn an outlier class to encapsulate the OOD samples. The main challenge is then to tackle the aforementioned confusion between OOD samples and head/tail-class samples when learning the outlier class. To this end, we introduce a novel calibrated outlier class learning (COCL) approach, in which 1) a debiased large margin learning method is introduced in the outlier class learning to distinguish OOD samples from both head and tail classes in the representation space and 2) an outlier-class-aware logit calibration method is defined to enhance the long-tailed classification confidence. Extensive empirical results on three popular benchmarks CIFAR10-LT, CIFAR100-LT, and ImageNet-LT demonstrate that COCL substantially outperforms state-of-the-art OOD detection methods in LTR while being able to improve the classification accuracy on ID data. Code is available at https://***/mala-lab/COCL. Copyright © 2023, The Authors. All rights reserved.

关键词： Image enhancement

Dense Network-Based Spectral-Polarization Image Fusion: Multispectral Data Enhancement via Encoder-Decoder Approach 23

学校读者我要写书评

暂无评论

Dense Network-Based Spectral-Polarization Image Fusion: Mult...

Proceedings of the 6th International Conference on Information Technologies and Electrical Engineering

作者： Shahid Karim Geng Tong Jinyang Li Yiting Yu Muhammad Ibrar Faisal Mehmood Research & Development Institute of Northwestern Polytechnical University in Shenzhen College of Mechanical Engineering Key Laboratory of Micro/Nano Systems for Aerospace Ministry of Education and Shaanxi Province Key Laboratory of Micro and Nano Electro-Mechanical Systems Northwestern Polytechnical University China Software College Shenyang Normal University China School of Electronic and Computer Engineering Peking University ShenZhen Graduate School China

ISBN: (纸本)9798400708299

Images are obtained by perceiving the sunlight reflected by objects or scenes, and due to limited solar irradiance, spatial resolution certainly decreases. In contrast, multispectral sensors hold the spatial information of ground objects even though the obtained images only have a few bands, which means that the obtained images have high spatial but insufficient spectral resolution. To overcome these limitations, we have developed a spectral and polarization image fusion technique that improves several perspectives of image fusion, such as intensity, edges, textures, and substantial information. Our proposed method utilizes an encoder-decoder approach, which contains dense and convolutional blocks to simulate the source images. The polarization images have been used to develop DoLP images that help to improve fused images’ quality. We can perceive the visual quality of fused images from the experimental results.

关键词： Dense module Image fusion Polarization images Spectral images

Dense Network-Based Spectral-Polarization Image Fusion: Multispectral Data Enhancement via Encoder-Decoder Approach 6

学校读者我要写书评

暂无评论

Dense Network-Based Spectral-Polarization Image Fusion: Mult...

6th International Conference on Information Technologies and Electrical Engineering, ICITEE 2023

作者： Karim, Shahid Tong, Geng Li, Jinyang Yu, Yiting Ibrar, Muhammad Mehmood, Faisal Research & Development Institute of Northwestern Polytechnical University in Shenzhen Shenzhen518057 China College of Mechanical Engineering Northwestern Polytechnical University Xi'an710072 China Key Laboratory of Micro/Nano Systems for Aerospace Ministry of Education Northwestern Polytechnical University Xi'an710072 China Shaanxi Province Key Laboratory of Micro and Nano Electro-Mechanical Systems Northwestern Polytechnical University Xi'an701172 China Research Center of Smart Sensing Chips Ningbo Institute of Northwestern Polytechnical University Ningbo315103 China Software College Shenyang Normal University Shenyang China School of Electronic and Computer Engineering Shenzhen Graduate School Peking University Shenzhen518055 China

ISBN: (纸本)9798400708299

Images are obtained by perceiving the sunlight reflected by objects or scenes, and due to limited solar irradiance, spatial resolution certainly decreases. In contrast, multispectral sensors hold the spatial information of ground objects even though the obtained images only have a few bands, which means that the obtained images have high spatial but insufficient spectral resolution. To overcome these limitations, we have developed a spectral and polarization image fusion technique that improves several perspectives of image fusion, such as intensity, edges, textures, and substantial information. Our proposed method utilizes an encoder-decoder approach, which contains dense and convolutional blocks to simulate the source images. The polarization images have been used to develop DoLP images that help to improve fused images' quality. We can perceive the visual quality of fused images from the experimental results. © 2023 ACM.

关键词： Image fusion

Unsupervised Recognition of Unknown Objects for Open-World Object Detection

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Fang, Ruohuan Pang, Guansong Zhou, Lei Bai, Xiao Zheng, Jin School of Computer Science and Engineering State Key Laboratory of Software Development Environment Jiangxi Research Institute Beihang University China School of Computing and Information Systems Singapore Management University Singapore School of Computer Science and Artificial Intelligence Wuhan University of Technology China

Open-World Object Detection (OWOD) extends object detection problem to a realistic and dynamic scenario, where a detection model is required to be capable of detecting both known and unknown objects and incrementally learning newly introduced knowledge. Current OWOD models, such as ORE and OW-DETR, focus on pseudo-labeling regions with high objectness scores as unknowns, whose performance relies heavily on the supervision of known objects. While they can detect the unknowns that exhibit similar features to the known objects, they suffer from a severe label bias problem that they tend to detect all regions (including unknown object regions) that are dissimilar to the known objects as part of the background. To eliminate the label bias, this paper proposes a novel approach that learns an unsupervised discriminative model to recognize true unknown objects from raw pseudo labels generated by unsupervised region proposal methods. The resulting model can be further refined by a classification-free self-training method which iteratively extends pseudo unknown objects to the unlabeled regions. Experimental results show that our method 1) significantly outperforms the prior SOTA in detecting unknown objects while maintaining competitive performance of detecting known object classes on the MS COCO dataset, and 2) achieves better generalization ability on the LVIS and Objects365 datasets. Code is available at https://***/frh23333/mepu-owod. Copyright © 2023, The Authors. All rights reserved.

关键词： Object detection

VS-Net: A Voxel Encoding and Sparse Convolution Embedded Network for LiDAR 3D Object Detection

学校读者我要写书评

暂无评论

VS-Net: A Voxel Encoding and Sparse Convolution Embedded Net...

2022 IEEE International Conference on Robotics and Biomimetics, ROBIO 2022

作者： Liu, Meng Niu, Jianwei Liu, Yu The China North Artificial Intelligence and Innovation Research Institute Beijing100072 China The China North Vehicle Research Institute Beijing100072 China The State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beihang University Beijing100191 China The State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing100191 China The Hangzhou Innovation Institute Beihang University Hangzhou310051 China The Zhengzhou University Research Institute of Industrial Technology School of Information Engineering Zhengzhou University Zhengzhou450001 China

ISBN: (纸本)9781665481090

Accurate and efficient 3D object detection in LiDAR point cloud is critical for autonomous driving. In this paper, we provide a solution for LiDAR-only detection which employs voxel features and a sparse convolution network. Specifically, we utilized voxel centers to encode point clouds' voxel features. Then, combining the sparse convolution method, we leveraged residual learning to design the backbone network. The detector output predicted results for multi-category and multi-object detection. Our method was evaluated on three popular and challenging datasets (KITTI, nuScence and Waymo). Experimental results demonstrated that the accuracy and speed (27.8 FPS) of our model could effectively support object detection for autonomous driving in various scenarios. © 2022 IEEE.

关键词： Object detection

Towards Label-Efficient Deep Learning-based Aging-related Bug Prediction with Spiking Convolutional Neural Networks

学校读者我要写书评

暂无评论

IEEE Transactions on Emerging Topics in Computing 2025年

作者： Tian, Yunzhe Li, Yike Chen, Kang Zhang, Zhenguo Tong, Endong Liu, Jiqiang Qin, Fangyun Zheng, Zheng Niu, Wenjia Hebei Boshilin Technology Development Co. Ltd. Shijiazhuang050051 China Beijing Jiaotong University Beijing Key Laboratory of Security and Privacy in Intelligent Transportation Beijing100044 China Tangshan Research Institute of Beijing Jiaotong University Tangshan063000 China Capital Normal University College of Information Engineering Beijing100048 China Nanjing University State Key Laboratory for Novel Software Technology Nanjing210093 China Beihang University Laboratory of Dependable Intelligent Systems School of Automation Science and Electrical Engineering Beijing100191 China

Recent advances in Deep Learning (DL) have enhanced Aging-Related Bug (ARB) prediction for mitigating software aging. However, DL-based ARB prediction models face a dual challenge: overcoming overfitting to enhance generalization and managing the high labeling costs associated with extensive data requirements. To address the first issue, we utilize the sparse and binary nature of spiking communication in Spiking Neural Networks (SNNs), which inherently provides brain-inspired regularization to effectively alleviate overfitting. Therefore, we propose a Spiking Convolutional Neural Network (SCNN)-based ARB prediction model along with a training framework that handles the model's spatial-temporal dynamics and non-differentiable nature. To reduce labeling costs, we introduce a Bio-inspired and Diversity-aware Active Learning framework (BiDAL), which prioritizes highly informative and diverse samples, enabling more efficient usage of the limited labeling budget. This framework incorporates bio-inspired uncertainty to enhance informativeness measurement along with using a diversity-aware selection strategy based on clustering to prevent redundant labeling. Experiments on three ARB datasets show that ARB-SCNN effectively reduces overfitting, improving generalization performance by 6.65% over other DL-based classifiers. Additionally, BiDAL boosts label efficiency for ARB-SCNN training, outperforming four state-of-the-art active learning methods by 4.77% within limited labeling budgets. © 2013 IEEE.

关键词： Active learning

VS-Net: A Voxel Encoding and Sparse Convolution Embedded Network for LiDAR 3D Object Detection

学校读者我要写书评

暂无评论

VS-Net: A Voxel Encoding and Sparse Convolution Embedded Net...

IEEE International Conference on Robotics and Biomimetics

作者： Meng Liu Jianwei Niu Yu Liu China North Artificial Intelligence and Innovation Research Institute Beijing China China North Vehicle Research Institute Beijing China State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beihang University Beijing China State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing China Hangzhou Innovation Institute Beihang University Hangzhou China Zhengzhou University Research Institute of Industrial Technology School of Information Engineering Zhengzhou University Zhengzhou China State Key Laboratory of Software Development Envi-ronment School of Computer Science and Engineering Beihang University Beijing China

ISBN: (纸本)9781665481106

关键词： Point cloud compression Laser radar Three-dimensional displays Convolution Object detection Detectors Feature extraction

Entity concept-enhanced few-shot relation extraction

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Yang, Shan Zhang, Yongfei Niu, Guanglin Zhao, Qinhua Pu, Shiliang Beijing Key Laboratory of Digital Media School of Computer Science and Engineering BeiHang University Beijing100191 China State Key Laboratory of Virtual Reality Technology and Systems BeiHang University Beijing100191 China Pengcheng Laboratory Shenzhen518055 China State Key Laboratory of Software Development Environment School of Computer Science and Engineering BeiHang University Beijing100191 China Hikvision Research Institute Hangzhou311500 China

Few-shot relation extraction (FSRE) is of great importance in long-tail distribution problem, especially in special domain with low-resource data. Most existing FSRE algorithms fail to accurately classify the relations merely based on the information of the sentences together with the recognized entity pairs, due to limited samples and lack of knowledge. To address this problem, in this paper, we proposed a novel entity CONCEPTenhanced FEw-shot Relation Extraction scheme (ConceptFERE), which introduces the inherent concepts of entities to provide clues for relation prediction and boost the relations classification performance. Firstly, a conceptsentence attention module is developed to select the most appropriate concept from multiple concepts of each entity by calculating the semantic similarity between sentences and concepts. Secondly, a self-attention based fusion module is presented to bridge the gap of concept embedding and sentence embedding from different semantic spaces. Extensive experiments on the FSRE benchmark dataset FewRel have demonstrated the effectiveness and the superiority of the proposed ConceptFERE scheme as compared to the state-of-the-art baselines. Code is available at https://***/LittleGuoKe/ConceptFERE. Copyright © 2021, The Authors. All rights reserved.

关键词： Embeddings