Recent works have shown that deep neural networks benefit from multi-task learning by learning a shared representation across several related tasks. However, performance of such systems depend on relative weighting between various losses involved during training. Prior works on loss weighting schemes assume that instances are equally easy or hard for all tasks. In order to break this assumption, we let the training process dictate the optimal weighting of tasks for every instance in the dataset. More specifically, we equip every instance in the dataset with a set of learnable parameters (instance-level task parameters) where the cardinality is equal to the number of tasks learned by the model. These parameters model the weighting of each task for an instance. They are updated by gradient descent and do not require hand-crafted rules. We conduct extensive experiments on SURREAL and CityScapes datasets, for human shape and pose estimation, depth estimation and semantic segmentation tasks. In these tasks, our approach outperforms recent dynamic loss weighting approaches, e.g. reducing surface estimation errors by 8.97% on SURREAL. When applied to datasets where one or more tasks can have noisy annotations, the proposed method learns to prioritize learning from clean labels for a given task, e.g. reducing surface estimation errors by up to 60%. We also show that we can reliably detect corrupt labels for a given task as a by-product from learned instance-level task parameters.
USIS-PGM: Photometric Gaussian Mixtures for Underwater Salient Instance Segmentation
Department of Electronic and Computer Engineering（电子与计算机工程系）；The Hong Kong University of Science and Technology（香港科学与技术大学）；School of CIT, Chair i6（CIT学院，Chair i6）；Technical University of Munich（慕尼黑技术大学）；School of CIT（CIT学院）；School of Robotics and Advanced Manufacture（机器人与先进制造学院）；Harbin Institute of Technology (Shenzhen)（哈尔滨工业大学（深圳））
Basque Center for Applied Mathematics (BCAM)（巴斯克应用数学中心（BCAM））；IKERBASQUE-Basque Foundation for Science（IKERBASQUE-巴斯克科学基金会）；Johns Hopkins University（约翰斯·霍普金斯大学）
Unsupervised Instance Segmentation in Microscopy Images via Panoptic Domain Adaptation and Task Re-weighting
University of Sydney（悉尼大学）；University of New South Wales（新南威尔士大学）；Brigham and Women’s Hospital（布莱根妇女医院）；Harvard Medical School（哈佛医学院）；University of Pittsburgh（匹兹堡大学）；Microsoft Corporation（微软公司）
Multiple Instance Learning with Auxiliary Task Weighting for Multiple Myeloma Classification
Imperial College London（帝国理工学院）；King’s College London（伦敦国王学院）；Imperial College Healthcare NHS Trust（帝国理工学院医疗保健国民保健服务信托基金）；The Institute of Cancer Research（癌症研究院）；The Royal Marsden NHS Foundation Trust（皇家马斯登国民保健服务基金会信托基金）
A Unified Framework for Unsupervised Domain Adaptation based on Instance Weighting
Southern University of Science and Technology（南方科技大学）；University of Technology Sydney（悉尼科技大学）；Eindhoven University of Technology（埃因霍温理工大学）；Peng Cheng Laboratory（鹏城实验室）；Hong Kong University of Science and Technology（香港科技大学）
Self-Training with Dynamic Weighting for Robust Gradual Domain Adaptation
University of Electronic Science and Technology of China（电子科技大学）；Tsinghua University（清华大学）；Zhenguan AI Lab；Southwest Jiaotong University（西南交通大学）；Shanghai University（上海大学）；New Jersey Institute of Technology（新泽西理工学院）
Coop-WD: Cooperative Perception with Weighting and Denoising for Robust V2V Communication
Department of Engineering, Durham University（工程系，达勒姆大学）；Faculty of Engineering and Information Technology, University of Technology, Sydney（工程与信息技术学院，悉尼技术大学）；Research Institute of Trustworthy Autonomous Systems, Southern University of Science and Technology（可信自主系统研究院，南方科技大学）
