Rethinking Performance Estimation in Neural Architecture Search
Xiamen University(厦门大学);National Institute for Data Science in Health and Medicine, Xiamen University(厦门大学健康医疗大数据国家研究院);Peng Cheng Laboratory(鹏城实验室);University of Chinese Academy of Sciences(中国科学院大学);Noah’s Ark Lab, Huawei Technologies(华为技术有限公司诺亚方舟实验室);Peking University(北京大学)
Xiawu Zheng, Rongrong Ji, Qiang Wang, Qixiang Ye, Zhenguo Li, Yonghong Tian, Qi Tian
Neural architecture search (NAS) remains a challenging problem, which is attributed to the indispensable and time-consuming component of performance estimation (PE). In this paper, we provide a novel yet systematic rethinking of PE in a resource constrained regime, termed budgeted PE (BPE), which precisely and effectively estimates the performance of an architecture sampled from an architecture space. Since searching an optimal BPE is extremely time-consuming as it requires to train a large number of networks for evaluation, we propose a Minimum Importance Pruning (MIP) approach. Given a dataset and a BPE search space, MIP estimates the importance of hyper-parameters using random forest and subsequently prunes the minimum one from the next iteration. In this way, MIP effectively prunes less important hyper-parameters to allocate more computational resource on more important ones, thus achieving an effective exploration. By combining BPE with various search algorithms including reinforcement learning, evolution algorithm, random search, and differentiable architecture search, we achieve 1, 000x of NAS speed up with a negligible performance drop comparing to the SOTA
Rethinking Unsupervised Cross-modal Flow Estimation: Learning from Decoupled Optimization and Consistency Constraint
College of Information Science and Electronic Engineering, Zhejiang University(浙江大学信息科学与电子工程学院);Ningbo Global Innovation Center, Zhejiang University(浙江大学宁波全球创新中心);NingboTech University(宁波科技大学)
Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所);University of Chinese Academy of Sciences(中国科学院大学);South China University of Technology(华南理工大学);Dalian Maritime University(大连海事大学)
Rethinking on Multi-Stage Networks for Human Pose Estimation
Megvii Inc. (Face++)(旷视科技(Face++));Shanghai Jiao Tong University(上海交通大学);Beihang University(北京航空航天大学);Beijing University of Posts and Telecommunications(北京邮电大学)
现有的姿态估计方法分为两类:单阶段和多阶段的方法。虽然多阶段方法似乎更适合这项任务,但在目前的实践中,它们的性能不如单阶段方法。本工作研究这一问题。我们认为,目前的多阶段方法的性能不令人满意的来源于各种设计选择的不足。我们提出了几个改进,包括单阶段的模块设计,跨阶段的功能聚合,和粗到细的监督。由此产生的方法在MS COCO和MPII Human Pose数据集上建立了新的最先进技术,证明了多阶段架构的有效性。源代码是公开的,可供进一步研究。
Rethinking the Heatmap Regression for Bottom-up Human Pose Estimation
MEGVII Research(旷视研究院);Megvii Inc(旷视科技);University of Chinese Academy of Sciences (UCAS)(中国科学院大学);Center for Research on Intelligent Perception and Computing (CRIPAC)(智能感知与计算研究中心);National Laboratory of Pattern Recognition (NLPR)(模式识别国家重点实验室);Institute of Automation, Chinese Academy of Sciences (CASIA)(中国科学院自动化研究所)
Wave2Body: Rethinking mmWave Human Pose Estimation as Radar-to-Body Token Translation
School of Computer Science, Peking University(北京大学计算机科学学院);University of Pittsburgh(匹兹堡大学);Key Laboratory of High Confidence Software Technologies, Ministry of Education (PKU)(教育部高可信软件技术重点实验室(北京大学))
Rethinking Keypoint Representations: Modeling Keypoints and Poses as Objects for Multi-Person Human Pose Estimation
University of Waterloo(滑铁卢大学);Waterloo Artificial Intelligence Institute(滑铁卢人工智能研究所)
在人体姿态估计等关键点估计任务中,基于热图的回归是主要方法,尽管具有显著的缺点:热图本质上存在量化误差,并且需要过多的计算来生成和后处理。为了找到更有效的解决方案,我们提出对各个关键点和空间相关关键点的集合(即,姿势)作为密集单级基于锚的检测框架内的对象。因此,我们将我们的方法称为KAPAO(发音为“Ka-Pow”),用于Keypoints And Poses As Objects。KAPAO适用于单阶段多人人体姿态估计的问题,同时检测人体姿态和关键点对象,并融合检测,以利用两个对象表示的优势。在实验中,我们观察到KAPAO比以前的方法更快,更准确,后者受到热图后处理的严重影响。当不使用测试时间增加时,在实际设置中,精度-速度的折衷是特别有利的。源代码:https://github.com/wmcnally/kapao。
Rethinking Skip Connections in Encoder-decoder Networks for Monocular Depth Estimation
Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences(中国科学院长春光学精密机械与物理研究所);University of the Chinese Academy of Sciences(中国科学院大学)
Rethinking Transparent Object Grasping: Depth Completion with Monocular Depth Estimation and Instance Mask
State Key Laboratory of Robotics and System at Harbin Institute of Technology(哈尔滨工业大学机器人系统国家重点实验室);Lanzhou University of Technology(兰州理工大学);Department of Computer Science(计算机科学系);University of Liverpool(利物浦大学)
在现有的基于学习的多视图立体视觉方法中,深度估计被解决为回归或分类问题。虽然这两种表示最近已经证明了它们的优异性能,但是它们仍然具有明显的缺点,例如,回归方法由于间接学习成本量而倾向于过拟合,并且分类方法由于其离散预测而不能直接推断精确深度。在本文中,我们提出了一种新的表示,称为统一,统一回归和分类的优点。它既可以像分类方法那样直接约束代价量,又可以像回归方法那样实现亚像素深度预测。为了挖掘统一性的潜力,我们设计了一个新的损失函数统一焦点损失,它更统一和合理,以应对样本不平衡的挑战。结合这两个无负担的模块,我们提出了一个由粗到细的框架,我们称之为UniMVSNet。在DTU和Tanks and Temples基准测试中排名第一的结果验证了我们的模型不仅性能最好,而且具有最好的泛化能力。
Action Machine: Rethinking Action Recognition in Trimmed Videos
Institute of Automation, Chinese Academy of Sciences (CASIA)(中国科学院自动化研究所);University of Chinese Academy of Sciences (UCAS)(中国科学院大学);Horizon Robotics, Inc.(地平线机器人公司);Xiamen University(厦门大学)
Rethinking Diffusion Posterior Sampling: From Conditional Score Estimator to Maximizing a Posterior
Institute for AI Industry Research, Tsinghua University(清华大学人工智能产业研究院);Department of Computer Science and Technology, Tsinghua University(清华大学计算机科学与技术系);New York University(纽约大学);Hong Kong University of Science and Technology(香港科学与技术大学);SenseTime Research(商汤科技研究院);The Chinese University of Hong Kong(香港中文大学);Kuaishou Technology(快手科技);Institute for Interdisciplinary Information Sciences, Tsinghua University(清华大学交叉信息研究院)
Reliable Deep Diffusion Tensor Estimation: Rethinking the Power of Data-Driven Optimization Routine
Southern Medical University(南方医科大学);JancsiTech;Nanjing University of Science and Technology(南京理工大学);Union Hospital, Tongji Medical College, Huazhong University of Science and Technology(华中科技大学同济医学院附属协和医院)