Pub Date : 2022-10-17DOI: 10.48550/arXiv.2210.09267
Jyh-Jing Hwang, Henrik Kretzschmar, Joshua M. Manela, Sean M. Rafferty, N. Armstrong-Crews, Tiffany Chen, Drago Anguelov
Robust 3D object detection is critical for safe autonomous driving. Camera and radar sensors are synergistic as they capture complementary information and work well under different environmental conditions. Fusing camera and radar data is challenging, however, as each of the sensors lacks information along a perpendicular axis, that is, depth is unknown to camera and elevation is unknown to radar. We propose the camera-radar matching network CramNet, an efficient approach to fuse the sensor readings from camera and radar in a joint 3D space. To leverage radar range measurements for better camera depth predictions, we propose a novel ray-constrained cross-attention mechanism that resolves the ambiguity in the geometric correspondences between camera features and radar features. Our method supports training with sensor modality dropout, which leads to robust 3D object detection, even when a camera or radar sensor suddenly malfunctions on a vehicle. We demonstrate the effectiveness of our fusion approach through extensive experiments on the RADIATE dataset, one of the few large-scale datasets that provide radar radio frequency imagery. A camera-only variant of our method achieves competitive performance in monocular 3D object detection on the Waymo Open Dataset.
{"title":"CramNet: Camera-Radar Fusion with Ray-Constrained Cross-Attention for Robust 3D Object Detection","authors":"Jyh-Jing Hwang, Henrik Kretzschmar, Joshua M. Manela, Sean M. Rafferty, N. Armstrong-Crews, Tiffany Chen, Drago Anguelov","doi":"10.48550/arXiv.2210.09267","DOIUrl":"https://doi.org/10.48550/arXiv.2210.09267","url":null,"abstract":"Robust 3D object detection is critical for safe autonomous driving. Camera and radar sensors are synergistic as they capture complementary information and work well under different environmental conditions. Fusing camera and radar data is challenging, however, as each of the sensors lacks information along a perpendicular axis, that is, depth is unknown to camera and elevation is unknown to radar. We propose the camera-radar matching network CramNet, an efficient approach to fuse the sensor readings from camera and radar in a joint 3D space. To leverage radar range measurements for better camera depth predictions, we propose a novel ray-constrained cross-attention mechanism that resolves the ambiguity in the geometric correspondences between camera features and radar features. Our method supports training with sensor modality dropout, which leads to robust 3D object detection, even when a camera or radar sensor suddenly malfunctions on a vehicle. We demonstrate the effectiveness of our fusion approach through extensive experiments on the RADIATE dataset, one of the few large-scale datasets that provide radar radio frequency imagery. A camera-only variant of our method achieves competitive performance in monocular 3D object detection on the Waymo Open Dataset.","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"11 1","pages":"388-405"},"PeriodicalIF":0.0,"publicationDate":"2022-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84939731","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-17DOI: 10.1007/978-3-031-20059-5_11
Sunjae Yoon, Jiajing Hong, Eunseop Yoon, Dahyun Kim, Junyeong Kim, Hee Suk Yoon, Changdong Yoo
{"title":"Selective Query-Guided Debiasing for Video Corpus Moment Retrieval","authors":"Sunjae Yoon, Jiajing Hong, Eunseop Yoon, Dahyun Kim, Junyeong Kim, Hee Suk Yoon, Changdong Yoo","doi":"10.1007/978-3-031-20059-5_11","DOIUrl":"https://doi.org/10.1007/978-3-031-20059-5_11","url":null,"abstract":"","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"1 1","pages":"185-200"},"PeriodicalIF":0.0,"publicationDate":"2022-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77177316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-17DOI: 10.48550/arXiv.2210.09022
Sanli Tang, Zhongyu Zhang, Zhanzhan Cheng, Jing Lu, Yunlu Xu, Yi Niu, Fan He
. Knowledge distillation learns a lightweight student model that mimics a cumbersome teacher. Existing methods regard the knowledge as the feature of each instance or their relations, which is the instance-level knowledge only from the teacher model, i.e., the local knowledge. However, the empirical studies show that the local knowledge is much noisy in object detection tasks, especially on the blurred, occluded, or small instances. Thus, a more intrinsic approach is to measure the representations of instances w.r.t. a group of common basis vectors in the two feature spaces of the teacher and the student detectors, i.e., global knowledge. Then, the distilling algorithm can be applied as space alignment. To this end, a novel prototype generation module (PGM) is proposed to find the common basis vectors, dubbed prototypes , in the two feature spaces. Then, a robust distilling module (RDM) is applied to construct the global knowledge based on the prototypes and filtrate noisy local knowledge by measuring the discrepancy of the representations in two feature spaces. Experiments with Faster-RCNN and RetinaNet on PASCAL and COCO datasets show that our method achieves the best performance for distilling object detectors with various backbones, which even surpasses the performance of the teacher model. We also show that the existing methods can be easily combined with global knowledge and obtain further improvement. Code is available: https://github.com/hikvision-research/DAVAR-Lab-ML . to (1) construct the global knowledge by projecting the instances w.r.t. the prototypes, and (2) robustly distill the global and local knowledge by measuring their discrepancy in the two spaces. Experiments show that the proposed method achieves state-of-the-art performance on two popular detection frameworks and benchmarks. The extensive experimental results show that the proposed method can be easily stretched with larger teachers and the existing knowledge distillation methods to obtain further improvement.
{"title":"Distilling Object Detectors With Global Knowledge","authors":"Sanli Tang, Zhongyu Zhang, Zhanzhan Cheng, Jing Lu, Yunlu Xu, Yi Niu, Fan He","doi":"10.48550/arXiv.2210.09022","DOIUrl":"https://doi.org/10.48550/arXiv.2210.09022","url":null,"abstract":". Knowledge distillation learns a lightweight student model that mimics a cumbersome teacher. Existing methods regard the knowledge as the feature of each instance or their relations, which is the instance-level knowledge only from the teacher model, i.e., the local knowledge. However, the empirical studies show that the local knowledge is much noisy in object detection tasks, especially on the blurred, occluded, or small instances. Thus, a more intrinsic approach is to measure the representations of instances w.r.t. a group of common basis vectors in the two feature spaces of the teacher and the student detectors, i.e., global knowledge. Then, the distilling algorithm can be applied as space alignment. To this end, a novel prototype generation module (PGM) is proposed to find the common basis vectors, dubbed prototypes , in the two feature spaces. Then, a robust distilling module (RDM) is applied to construct the global knowledge based on the prototypes and filtrate noisy local knowledge by measuring the discrepancy of the representations in two feature spaces. Experiments with Faster-RCNN and RetinaNet on PASCAL and COCO datasets show that our method achieves the best performance for distilling object detectors with various backbones, which even surpasses the performance of the teacher model. We also show that the existing methods can be easily combined with global knowledge and obtain further improvement. Code is available: https://github.com/hikvision-research/DAVAR-Lab-ML . to (1) construct the global knowledge by projecting the instances w.r.t. the prototypes, and (2) robustly distill the global and local knowledge by measuring their discrepancy in the two spaces. Experiments show that the proposed method achieves state-of-the-art performance on two popular detection frameworks and benchmarks. The extensive experimental results show that the proposed method can be easily stretched with larger teachers and the existing knowledge distillation methods to obtain further improvement.","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"18 1","pages":"422-438"},"PeriodicalIF":0.0,"publicationDate":"2022-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81107469","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-15DOI: 10.48550/arXiv.2210.08161
Hao Feng, Wen-gang Zhou, Jiajun Deng, Yuechen Wang, Houqiang Li
In document image rectification, there exist rich geometric constraints between the distorted image and the ground truth one. However, such geometric constraints are largely ignored in existing advanced solutions, which limits the rectification performance. To this end, we present DocGeoNet for document image rectification by introducing explicit geometric representation. Technically, two typical attributes of the document image are involved in the proposed geometric representation learning, i.e., 3D shape and textlines. Our motivation arises from the insight that 3D shape provides global unwarping cues for rectifying a distorted document image while overlooking the local structure. On the other hand, textlines complementarily provide explicit geometric constraints for local patterns. The learned geometric representation effectively bridges the distorted image and the ground truth one. Extensive experiments show the effectiveness of our framework and demonstrate the superiority of our DocGeoNet over state-of-the-art methods on both the DocUNet Benchmark dataset and our proposed DIR300 test set. The code is available at https://github.com/fh2019ustc/DocGeoNet.
{"title":"Geometric Representation Learning for Document Image Rectification","authors":"Hao Feng, Wen-gang Zhou, Jiajun Deng, Yuechen Wang, Houqiang Li","doi":"10.48550/arXiv.2210.08161","DOIUrl":"https://doi.org/10.48550/arXiv.2210.08161","url":null,"abstract":"In document image rectification, there exist rich geometric constraints between the distorted image and the ground truth one. However, such geometric constraints are largely ignored in existing advanced solutions, which limits the rectification performance. To this end, we present DocGeoNet for document image rectification by introducing explicit geometric representation. Technically, two typical attributes of the document image are involved in the proposed geometric representation learning, i.e., 3D shape and textlines. Our motivation arises from the insight that 3D shape provides global unwarping cues for rectifying a distorted document image while overlooking the local structure. On the other hand, textlines complementarily provide explicit geometric constraints for local patterns. The learned geometric representation effectively bridges the distorted image and the ground truth one. Extensive experiments show the effectiveness of our framework and demonstrate the superiority of our DocGeoNet over state-of-the-art methods on both the DocUNet Benchmark dataset and our proposed DIR300 test set. The code is available at https://github.com/fh2019ustc/DocGeoNet.","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"607 1","pages":"475-492"},"PeriodicalIF":0.0,"publicationDate":"2022-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78955782","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-14DOI: 10.48550/arXiv.2210.08064
Minghua Liu, Yin Zhou, C. Qi, Boqing Gong, Hao Su, Drago Anguelov
Semantic segmentation of LiDAR point clouds is an important task in autonomous driving. However, training deep models via conventional supervised methods requires large datasets which are costly to label. It is critical to have label-efficient segmentation approaches to scale up the model to new operational domains or to improve performance on rare cases. While most prior works focus on indoor scenes, we are one of the first to propose a label-efficient semantic segmentation pipeline for outdoor scenes with LiDAR point clouds. Our method co-designs an efficient labeling process with semi/weakly supervised learning and is applicable to nearly any 3D semantic segmentation backbones. Specifically, we leverage geometry patterns in outdoor scenes to have a heuristic pre-segmentation to reduce the manual labeling and jointly design the learning targets with the labeling process. In the learning step, we leverage prototype learning to get more descriptive point embeddings and use multi-scan distillation to exploit richer semantics from temporally aggregated point clouds to boost the performance of single-scan models. Evaluated on the SemanticKITTI and the nuScenes datasets, we show that our proposed method outperforms existing label-efficient methods. With extremely limited human annotations (e.g., 0.1% point labels), our proposed method is even highly competitive compared to the fully supervised counterpart with 100% labels.
{"title":"LESS: Label-Efficient Semantic Segmentation for LiDAR Point Clouds","authors":"Minghua Liu, Yin Zhou, C. Qi, Boqing Gong, Hao Su, Drago Anguelov","doi":"10.48550/arXiv.2210.08064","DOIUrl":"https://doi.org/10.48550/arXiv.2210.08064","url":null,"abstract":"Semantic segmentation of LiDAR point clouds is an important task in autonomous driving. However, training deep models via conventional supervised methods requires large datasets which are costly to label. It is critical to have label-efficient segmentation approaches to scale up the model to new operational domains or to improve performance on rare cases. While most prior works focus on indoor scenes, we are one of the first to propose a label-efficient semantic segmentation pipeline for outdoor scenes with LiDAR point clouds. Our method co-designs an efficient labeling process with semi/weakly supervised learning and is applicable to nearly any 3D semantic segmentation backbones. Specifically, we leverage geometry patterns in outdoor scenes to have a heuristic pre-segmentation to reduce the manual labeling and jointly design the learning targets with the labeling process. In the learning step, we leverage prototype learning to get more descriptive point embeddings and use multi-scan distillation to exploit richer semantics from temporally aggregated point clouds to boost the performance of single-scan models. Evaluated on the SemanticKITTI and the nuScenes datasets, we show that our proposed method outperforms existing label-efficient methods. With extremely limited human annotations (e.g., 0.1% point labels), our proposed method is even highly competitive compared to the fully supervised counterpart with 100% labels.","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"75 1","pages":"70-89"},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73860713","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-14DOI: 10.48550/arXiv.2210.07489
Hyeonsu Lee, Chankyu Choi
Scene text removal (STR), a task of erasing text from natural scene images, has recently attracted attention as an important component of editing text or concealing private information such as ID, telephone, and license plate numbers. While there are a variety of different methods for STR actively being researched, it is difficult to evaluate superiority because previously proposed methods do not use the same standardized training/evaluation dataset. We use the same standardized training/testing dataset to evaluate the performance of several previous methods after standardized re-implementation. We also introduce a simple yet extremely effective Gated Attention (GA) and Region-of-Interest Generation (RoIG) methodology in this paper. GA uses attention to focus on the text stroke as well as the textures and colors of the surrounding regions to remove text from the input image much more precisely. RoIG is applied to focus on only the region with text instead of the entire image to train the model more efficiently. Experimental results on the benchmark dataset show that our method significantly outperforms existing state-of-the-art methods in almost all metrics with remarkably higher-quality results. Furthermore, because our model does not generate a text stroke mask explicitly, there is no need for additional refinement steps or sub-models, making our model extremely fast with fewer parameters. The dataset and code are available at this https://github.com/naver/garnet.
场景文本删除(Scene text removal, STR)是一种从自然场景图像中删除文本的技术，近年来作为编辑文本或隐藏身份证件、电话号码、车牌号码等私人信息的重要组成部分而受到关注。虽然有各种不同的STR方法正在被积极研究，但很难评估其优越性，因为之前提出的方法没有使用相同的标准化训练/评估数据集。我们使用相同的标准化训练/测试数据集来评估标准化后重新实现的几种先前方法的性能。本文还介绍了一种简单但非常有效的门控注意(GA)和兴趣区域生成(RoIG)方法。GA将注意力集中在文本笔画以及周围区域的纹理和颜色上，从而更精确地从输入图像中删除文本。RoIG应用于只关注有文本的区域，而不是整个图像，以更有效地训练模型。在基准数据集上的实验结果表明，我们的方法在几乎所有指标上都明显优于现有的最先进的方法，并且结果质量显著提高。此外，由于我们的模型不显式地生成文本笔画遮罩，因此不需要额外的细化步骤或子模型，从而使我们的模型以更少的参数非常快。数据集和代码可在此获得https://github.com/naver/garnet。
{"title":"The Surprisingly Straightforward Scene Text Removal Method With Gated Attention and Region of Interest Generation: A Comprehensive Prominent Model Analysis","authors":"Hyeonsu Lee, Chankyu Choi","doi":"10.48550/arXiv.2210.07489","DOIUrl":"https://doi.org/10.48550/arXiv.2210.07489","url":null,"abstract":"Scene text removal (STR), a task of erasing text from natural scene images, has recently attracted attention as an important component of editing text or concealing private information such as ID, telephone, and license plate numbers. While there are a variety of different methods for STR actively being researched, it is difficult to evaluate superiority because previously proposed methods do not use the same standardized training/evaluation dataset. We use the same standardized training/testing dataset to evaluate the performance of several previous methods after standardized re-implementation. We also introduce a simple yet extremely effective Gated Attention (GA) and Region-of-Interest Generation (RoIG) methodology in this paper. GA uses attention to focus on the text stroke as well as the textures and colors of the surrounding regions to remove text from the input image much more precisely. RoIG is applied to focus on only the region with text instead of the entire image to train the model more efficiently. Experimental results on the benchmark dataset show that our method significantly outperforms existing state-of-the-art methods in almost all metrics with remarkably higher-quality results. Furthermore, because our model does not generate a text stroke mask explicitly, there is no need for additional refinement steps or sub-models, making our model extremely fast with fewer parameters. The dataset and code are available at this https://github.com/naver/garnet.","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"31 1","pages":"457-472"},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85301718","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-14DOI: 10.48550/arXiv.2210.08061
Mahyar Najibi, Jingwei Ji, Yin Zhou, C. Qi, Xinchen Yan, S. Ettinger, Drago Anguelov
Learning-based perception and prediction modules in modern autonomous driving systems typically rely on expensive human annotation and are designed to perceive only a handful of predefined object categories. This closed-set paradigm is insufficient for the safety-critical autonomous driving task, where the autonomous vehicle needs to process arbitrarily many types of traffic participants and their motion behaviors in a highly dynamic world. To address this difficulty, this paper pioneers a novel and challenging direction, i.e., training perception and prediction models to understand open-set moving objects, with no human supervision. Our proposed framework uses self-learned flow to trigger an automated meta labeling pipeline to achieve automatic supervision. 3D detection experiments on the Waymo Open Dataset show that our method significantly outperforms classical unsupervised approaches and is even competitive to the counterpart with supervised scene flow. We further show that our approach generates highly promising results in open-set 3D detection and trajectory prediction, confirming its potential in closing the safety gap of fully supervised systems.
{"title":"Motion Inspired Unsupervised Perception and Prediction in Autonomous Driving","authors":"Mahyar Najibi, Jingwei Ji, Yin Zhou, C. Qi, Xinchen Yan, S. Ettinger, Drago Anguelov","doi":"10.48550/arXiv.2210.08061","DOIUrl":"https://doi.org/10.48550/arXiv.2210.08061","url":null,"abstract":"Learning-based perception and prediction modules in modern autonomous driving systems typically rely on expensive human annotation and are designed to perceive only a handful of predefined object categories. This closed-set paradigm is insufficient for the safety-critical autonomous driving task, where the autonomous vehicle needs to process arbitrarily many types of traffic participants and their motion behaviors in a highly dynamic world. To address this difficulty, this paper pioneers a novel and challenging direction, i.e., training perception and prediction models to understand open-set moving objects, with no human supervision. Our proposed framework uses self-learned flow to trigger an automated meta labeling pipeline to achieve automatic supervision. 3D detection experiments on the Waymo Open Dataset show that our method significantly outperforms classical unsupervised approaches and is even competitive to the counterpart with supervised scene flow. We further show that our approach generates highly promising results in open-set 3D detection and trajectory prediction, confirming its potential in closing the safety gap of fully supervised systems.","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"102 1","pages":"424-443"},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75645026","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.48550/arXiv.2210.06686
Yu-Zhu Guo, Liwen Zhang, Y. Chen, Xinyi Tong, Xiaode Liu, Yinglei Wang, Xuhui Huang, Zhe Ma
Brain-inspired spiking neural networks (SNNs) have recently drawn more and more attention due to their event-driven and energyefficient characteristics. The integration of storage and computation paradigm on neuromorphic hardwares makes SNNs much different from Deep Neural Networks (DNNs). In this paper, we argue that SNNs may not benefit from the weight-sharing mechanism, which can effectively reduce parameters and improve inference efficiency in DNNs, in some hardwares, and assume that an SNN with unshared convolution kernels could perform better. Motivated by this assumption, a training-inference decoupling method for SNNs named as Real Spike is proposed, which not only enjoys both unshared convolution kernels and binary spikes in inference-time but also maintains both shared convolution kernels and Real-valued Spikes during training. This decoupling mechanism of SNN is realized by a re-parameterization technique. Furthermore, based on the training-inference-decoupled idea, a series of different forms for implementing Real Spike on different levels are presented, which also enjoy shared convolutions in the inference and are friendly to both neuromorphic and non-neuromorphic hardware platforms. A theoretical proof is given to clarify that the Real Spike-based SNN network is superior to its vanilla counterpart. Experimental results show that all different Real Spike versions can consistently improve the SNN performance. Moreover, the proposed method outperforms the state-of-the-art models on both non-spiking static and neuromorphic datasets.
{"title":"Real Spike: Learning Real-valued Spikes for Spiking Neural Networks","authors":"Yu-Zhu Guo, Liwen Zhang, Y. Chen, Xinyi Tong, Xiaode Liu, Yinglei Wang, Xuhui Huang, Zhe Ma","doi":"10.48550/arXiv.2210.06686","DOIUrl":"https://doi.org/10.48550/arXiv.2210.06686","url":null,"abstract":"Brain-inspired spiking neural networks (SNNs) have recently drawn more and more attention due to their event-driven and energyefficient characteristics. The integration of storage and computation paradigm on neuromorphic hardwares makes SNNs much different from Deep Neural Networks (DNNs). In this paper, we argue that SNNs may not benefit from the weight-sharing mechanism, which can effectively reduce parameters and improve inference efficiency in DNNs, in some hardwares, and assume that an SNN with unshared convolution kernels could perform better. Motivated by this assumption, a training-inference decoupling method for SNNs named as Real Spike is proposed, which not only enjoys both unshared convolution kernels and binary spikes in inference-time but also maintains both shared convolution kernels and Real-valued Spikes during training. This decoupling mechanism of SNN is realized by a re-parameterization technique. Furthermore, based on the training-inference-decoupled idea, a series of different forms for implementing Real Spike on different levels are presented, which also enjoy shared convolutions in the inference and are friendly to both neuromorphic and non-neuromorphic hardware platforms. A theoretical proof is given to clarify that the Real Spike-based SNN network is superior to its vanilla counterpart. Experimental results show that all different Real Spike versions can consistently improve the SNN performance. Moreover, the proposed method outperforms the state-of-the-art models on both non-spiking static and neuromorphic datasets.","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"108 1","pages":"52-68"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86125613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-13DOI: 10.48550/arXiv.2210.07424
Yuxuan Liu, Nikhil Mishra, Maximilian Sieb, Yide Shentu, P. Abbeel, Xi Chen
. 3D bounding boxes are a widespread intermediate representation in many computer vision applications. However, predicting them is a challenging task, largely due to partial observability, which motivates the need for a strong sense of uncertainty. While many recent methods have explored better architectures for consuming sparse and unstructured point cloud data, we hypothesize that there is room for improve-ment in the modeling of the output distribution and explore how this can be achieved using an autoregressive prediction head. Additionally, we release a simulated dataset, COB-3D, which highlights new types of ambiguity that arise in real-world robotics applications, where 3D bounding box prediction has largely been underexplored. We propose methods for leveraging our autoregressive model to make high confidence predictions and meaningful uncertainty measures, achieving strong results on SUN-RGBD, Scannet, KITTI, and our new dataset 3 .
{"title":"Autoregressive Uncertainty Modeling for 3D Bounding Box Prediction","authors":"Yuxuan Liu, Nikhil Mishra, Maximilian Sieb, Yide Shentu, P. Abbeel, Xi Chen","doi":"10.48550/arXiv.2210.07424","DOIUrl":"https://doi.org/10.48550/arXiv.2210.07424","url":null,"abstract":". 3D bounding boxes are a widespread intermediate representation in many computer vision applications. However, predicting them is a challenging task, largely due to partial observability, which motivates the need for a strong sense of uncertainty. While many recent methods have explored better architectures for consuming sparse and unstructured point cloud data, we hypothesize that there is room for improve-ment in the modeling of the output distribution and explore how this can be achieved using an autoregressive prediction head. Additionally, we release a simulated dataset, COB-3D, which highlights new types of ambiguity that arise in real-world robotics applications, where 3D bounding box prediction has largely been underexplored. We propose methods for leveraging our autoregressive model to make high confidence predictions and meaningful uncertainty measures, achieving strong results on SUN-RGBD, Scannet, KITTI, and our new dataset 3 .","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"33 1","pages":"673-694"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91030375","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
3D object detection in point clouds is a core component for modern robotics and autonomous driving systems. A key challenge in 3D object detection comes from the inherent sparse nature of point occupancy within the 3D scene. In this paper, we propose Sparse Window Transformer (SWFormer ), a scalable and accurate model for 3D object detection, which can take full advantage of the sparsity of point clouds. Built upon the idea of window-based Transformers, SWFormer converts 3D points into sparse voxels and windows, and then processes these variable-length sparse windows efficiently using a bucketing scheme. In addition to self-attention within each spatial window, our SWFormer also captures cross-window correlation with multi-scale feature fusion and window shifting operations. To further address the unique challenge of detecting 3D objects accurately from sparse features, we propose a new voxel diffusion technique. Experimental results on the Waymo Open Dataset show our SWFormer achieves state-of-the-art 73.36 L2 mAPH on vehicle and pedestrian for 3D object detection on the official test set, outperforming all previous single-stage and two-stage models, while being much more efficient.
{"title":"SWFormer: Sparse Window Transformer for 3D Object Detection in Point Clouds","authors":"Pei Sun, Mingxing Tan, Weiyue Wang, Chenxi Liu, Fei Xia, Zhaoqi Leng, Drago Anguelov","doi":"10.48550/arXiv.2210.07372","DOIUrl":"https://doi.org/10.48550/arXiv.2210.07372","url":null,"abstract":"3D object detection in point clouds is a core component for modern robotics and autonomous driving systems. A key challenge in 3D object detection comes from the inherent sparse nature of point occupancy within the 3D scene. In this paper, we propose Sparse Window Transformer (SWFormer ), a scalable and accurate model for 3D object detection, which can take full advantage of the sparsity of point clouds. Built upon the idea of window-based Transformers, SWFormer converts 3D points into sparse voxels and windows, and then processes these variable-length sparse windows efficiently using a bucketing scheme. In addition to self-attention within each spatial window, our SWFormer also captures cross-window correlation with multi-scale feature fusion and window shifting operations. To further address the unique challenge of detecting 3D objects accurately from sparse features, we propose a new voxel diffusion technique. Experimental results on the Waymo Open Dataset show our SWFormer achieves state-of-the-art 73.36 L2 mAPH on vehicle and pedestrian for 3D object detection on the official test set, outperforming all previous single-stage and two-stage models, while being much more efficient.","PeriodicalId":72676,"journal":{"name":"Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision","volume":"119 17","pages":"426-442"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91408690","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：