检索结果-内蒙古大学图书馆

Sim-to-Real Grasp Detection with Global-to-Local RGB-D Adaptation

学校读者我要写书评

暂无评论

Sim-to-Real Grasp Detection with Global-to-Local RGB-D Adapt...

IEEE International Conference on Robotics and Automation (ICRA)

作者： Haoxiang Ma Ran Qin Modi Shi Boyang Gao Di Huang State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beihang University Beijing China Geometry Robotics and the School of Computer Science and Technology Harbin Institute of Technology Harbin China

ISBN: (数字)9798350384574

ISBN: (纸本)9798350384581

This paper focuses on the sim-to-real issue of RGB-D grasp detection and formulates it as a domain adaptation problem. In this case, we present a global-to-local method to address hybrid domain gaps in RGB and depth data and insufficient multi-modal feature alignment. First, a self-supervised rotation pre-training strategy is adopted to deliver robust initialization for RGB and depth networks. We then propose a global-to-local alignment pipeline with individual global domain classifiers for scene features of RGB and depth images as well as a local one specifically working for grasp features in the two modalities. In particular, we propose a grasp prototype adaptation module, which aims to facilitate fine-grained local feature alignment by dynamically updating and matching the grasp prototypes from the simulation and real-world scenarios throughout the training process. Due to such designs, the proposed method substantially reduces the domain shift and thus leads to consistent performance improvements. Extensive experiments are conducted on the GraspNet-Planar benchmark and physical environment, and superior results are achieved which demonstrate the effectiveness of our method. Code is available at https://***/mahaoxiang822/GL-MSDA.

关键词： Training Codes Pipelines Prototypes Benchmark testing Robustness Robotics and automation

CROP: Zero-shot Cross-lingual Named Entity Recognition with Multilingual labeled Sequence Translation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yang, Jian Huang, Shaohan Ma, Shuming Yin, Yuwei Dong, Li Zhang, Dongdong Guo, Hongcheng Li, Zhoujun Wei, Furu State Key Lab of Software Development Environment Beihang University China Microsoft Research Asia China The University of Hong Kong Hong Kong

Named entity recognition (NER) suffers from the scarcity of annotated training data, especially for low-resource languages without labeled data. Cross-lingual NER has been proposed to alleviate this issue by transferring knowledge from high-resource languages to low-resource languages via aligned cross-lingual representations or machine translation results. However, the performance of cross-lingual NER methods is severely affected by the unsatisfactory quality of translation or label projection. To address these problems, we propose a Cross-lingual Entity Projection framework (CROP) to enable zero-shot cross-lingual NER with the help of a multilingual labeled sequence translation model. Specifically, the target sequence is first translated into the source language and then tagged by a source NER model. We further adopt a labeled sequence translation model to project the tagged sequence back to the target language and label the target raw sentence. Ultimately, the whole pipeline is integrated into an end-to-end model by the way of self-training. Experimental results on two benchmarks demonstrate that our method substantially outperforms the previous strong baseline by a large margin of +3∼7 F1 scores and achieves state-of-the-art performance. Copyright © 2022, The Authors. All rights reserved.

关键词： Benchmarking

Centerness-based Instance-aware Knowledge Distillation with Task-wise Mutual Lifting for Object Detection on Drone Imagery

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Du, Bowei Liao, Zhixuan Zhang, Yanan Cai, Zhi Chen, Jiaxin Huang, Di State Key Laboratory of Software Development Environment Beihang University Beijing China School of Computer Science and Engineering Beihang University Beijing China Hangzhou Innovation Institute Beihang University Hangzhou China

Developing accurate and efficient detectors for drone imagery is challenging due to the inherent complexity of aerial scenes. While some existing methods aim to achieve high accuracy by utilizing larger models, their computational cost is prohibitive for drones. Recently, Knowledge Distillation (KD) has shown promising potential for maintaining satisfactory accuracy while significantly compressing models in general object detection. Considering the advantages of KD, this paper presents the first attempt to adapt it to object detection on drone imagery and addresses two intrinsic issues: (1) low foreground-background ratio and (2) small instances and complex backgrounds, which lead to inadequate training, resulting insufficient distillation. Therefore, we propose a task-wise Lightweight Mutual Lifting (Light-ML) module with a Centerness-based Instance-aware Distillation (CID) strategy. The Light-ML module mutually harmonizes the classification and localization branches by channel shuffling and convolution, integrating teacher supervision across different tasks during back-propagation, thus facilitating training the student model. The CID strategy extracts valuable regions surrounding instances through the centerness of proposals, enhancing distillation efficacy. Experiments on the VisDrone, UAVDT, and COCO benchmarks demonstrate that the proposed approach promotes the accuracies of existing state-ofthe- art KD methods with comparable computational requirements. Codes will be available upon acceptance. Copyright © 2024, The Authors. All rights reserved.

关键词： Drones

Sim-to-Real Grasp Detection with Global-to-Local RGB-D Adaptation

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Ma, Haoxiang Qin, Ran Shi, Modi Gao, Boyang Huang, Di The State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beihang University Beijing China The Geometry Robotics and the School of Computer Science and Technology Harbin Institute of Technology Harbin China

关键词：

PAEG: Phrase-level Adversarial Example Generation for Neural Machine Translation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Wan, Juncheng Yang, Jian Ma, Shuming Zhang, Dongdong Zhang, Weinan Yu, Yong Li, Zhoujun Shanghai Jiao Tong University China State Key Lab of Software Development Environment Beihang University China Microsoft Research Asia China

While end-to-end neural machine translation (NMT) has achieved impressive progress, noisy input usually leads models to become fragile and unstable. Generating adversarial examples as the augmented data has been proved to be useful to alleviate this problem. Existing methods for adversarial example generation (AEG) are word-level or character-level, which ignore the ubiquitous phrase structure. In this paper, we propose a Phrase-level Adversarial Example Generation (PAEG) framework to enhance the robustness of the translation model. Our method further improves the gradient-based word-level AEG method by adopting a phrase-level substitution strategy. We verify our method on three benchmarks, including LDC Chinese-English, IWSLT14 German-English, and WMT14 English-German tasks. Experimental results demonstrate that our approach significantly improves translation performance and robustness to noise compared to previous strong baselines. Copyright © 2022, The Authors. All rights reserved.

关键词： Neural machine translation

GANLM: Encoder-Decoder Pre-training with an Auxiliary Discriminator

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yang, Jian Ma, Shuming Dong, Li Huang, Shaohan Huang, Haoyang Yin, Yuwei Zhang, Dongdong Yang, Liqun Wei, Furu Li, Zhoujun State Key Lab of Software Development Environment Beihang University China Microsoft Research Asia China The University of Hong Kong Hong Kong

Pre-trained models have achieved remarkable success in natural language processing (NLP). However, existing pre-training methods under-utilize the benefits of language understanding for generation. Inspired by the idea of Generative Adversarial Networks (GANs), we propose a GAN-style model for encoder-decoder pre-training by introducing an auxiliary discriminator, unifying the ability of language understanding and generation in a single model. Our model, named as GANLM, is trained with two pre-training objectives: replaced token detection and replaced token denoising. Specifically, given masked source sentences, the generator outputs the target distribution and the discriminator predicts whether the target sampled tokens from distribution are incorrect. The target sentence is replaced with misclassified tokens to construct noisy previous context, which is used to generate the gold sentence. In general, both tasks improve the ability of language understanding and generation by selectively using the denoising data. Extensive experiments in language generation benchmarks show that GANLM with the powerful language understanding capability outperforms various strong pre-trained language models (PLMs) and achieves state-of-the-art performance.1 Copyright © 2022, The Authors. All rights reserved.

关键词： Discriminators

Relative probabilities of breakup channels in reactions of^(6.7)Li with^(209)Biat energies around and above the Coulomb barrier

学校读者我要写书评

暂无评论

Chinese Physics C 2021年第5期45卷 141-148页

作者： Yong-Jin Yao Cheng-Jian Lin Lei Yang Nan-Ru Ma Dong-Xi Wang Gao-long Zhang Guang-Xin Zhang Hui-Ming Jia Feng Yang Fu-Peng Zhong Pei-Wei Wen Xiu-Bo Qin Hong-Ming Zhao State Key Laboratory of Software Development Environment Beihang UniversityBeijing 100191China School of Physics Beihang UniversityBejing 100191China China Institute of Atomic Energy Beijing 102413China Guangxi Normal University Guangxi 541004China X Lab the Second Academy of CASICBejjing 100854China

Coincidence measurements of breakup fragments in reactions of^(6.7)with^(209)Bi at energies around and above the Coulomb barrier were carried out using a large solid-angle covered detector *** the Q values along with the relative energies of the breakup fragments,different breakup components(prompt breakups and delayed breakups)and different breakup modes(α+t,α+d,α+p,andα+α)are distinguished.A new breakup mode,α+t,is observed in^(6)Li-induced reactions at energies above the Coulomb *** between breakup modes and breakup components as well as their variations with the incident energy are *** results will help us better understand the breakup effects of weakly bound nuclei on the suppression of a complete fusion,particularly for the above-barrier energies.

关键词： breakup mechanism coincidence measurement weakly bound nucleus

FVP: Fourier Visual Prompting for Source-Free Unsupervised Domain Adaptation of Medical Image Segmentation

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Yan Cheng, Jian Chen, Yixin Shao, Shuai Zhu, Lanyun Wu, Zhenzhou Liu, Tao Zhu, Haogang School of Instrumentation and Optoelectronic Engineering State Key Laboratory of Software Development Environment Beihang University Beijing100191 China Zhejiang Lab Hangzhou311100 China pillar Singapore University of Technology and Design Singapore The BioMind Technology Center Beijing China The School of Biological Science and Medical Engineering Beihang University China State Key Laboratory of Software Development Environment Beihang University Beijing100191 China Zhongguancun Laboratory China

Medical image segmentation methods normally perform poorly when there is a domain shift between training and testing data. Unsupervised Domain Adaptation (UDA) addresses the domain shift problem by training the model using both labeled data from the source domain and unlabeled data from the target domain. Source-Free UDA (SFUDA) was recently proposed for UDA without requiring the source data during the adaptation, due to data privacy or data transmission issues, which normally adapts the pre-trained deep model in the testing stage. However, in real clinical scenarios of medical image segmentation, the trained model is normally frozen in the testing stage. In this paper, we propose Fourier Visual Prompting (FVP) for SFUDA of medical image segmentation. Inspired by prompting learning in natural language processing, FVP steers the frozen pre-trained model to perform well in the target domain by adding a visual prompt to the input target data. In FVP, the visual prompt is parameterized using only a small amount of low-frequency learnable parameters in the input frequency space, and is learned by minimizing the segmentation loss between the predicted segmentation of the prompted target image and reliable pseudo segmentation label of the target image under the frozen model. To our knowledge, FVP is the first work to apply visual prompts to SFUDA for medical image segmentation. The proposed FVP is validated using three public datasets, and experiments demonstrate that FVP yields better segmentation results, compared with various existing methods. Copyright © 2023, The Authors. All rights reserved.

关键词： Image segmentation

BIPOINTNET: BINARY NEURAL NETWORK FOR POINT CLOUDS 9

学校读者我要写书评

暂无评论

BIPOINTNET: BINARY NEURAL NETWORK FOR POINT CLOUDS

9th International Conference on Learning Representations, ICLR 2021

作者： Qin, Haotong Cai, Zhongang Zhang, Mingyuan Ding, Yifu Zhao, Haiyu Yi, Shuai Liu, Xianglong Su, Hao State Key Lab of Software Development Environment Beihang University China Shen Yuan Honors College Beihang University China SenseTime Research University of California San Diego United States

To alleviate the resource constraint for real-time point cloud applications that run on edge devices, in this paper we present BiPointNet, the first model binarization approach for efficient deep learning on point clouds. We discover that the immense performance drop of binarized models for point clouds mainly stems from two challenges: aggregation-induced feature homogenization that leads to a degradation of information entropy, and scale distortion that hinders optimization and invalidates scale-sensitive structures. With theoretical justifications and in-depth analysis, our BiPointNet introduces Entropy-Maximizing Aggregation (EMA) to modulate the distribution before aggregation for the maximum information entropy, and Layer-wise Scale Recovery (LSR) to efficiently restore feature representation capacity. Extensive experiments show that BiPointNet outperforms existing binarization methods by convincing margins, at the level even comparable with the full precision counterpart. We highlight that our techniques are generic, guaranteeing significant improvements on various fundamental tasks and mainstream backbones. Moreover, BiPointNet gives an impressive 14.7× speedup and 18.9× storage saving on real-world resource-constrained devices. © 2021 ICLR 2021 - 9th International Conference on Learning Representations. All rights reserved.

关键词： Deep learning