第四色站

  • <tr id='njDbst'><strong id='njDbst'></strong><small id='njDbst'></small><button id='njDbst'></button><li id='njDbst'><noscript id='njDbst'><big id='njDbst'></big><dt id='njDbst'></dt></noscript></li></tr><ol id='njDbst'><option id='njDbst'><table id='njDbst'><blockquote id='njDbst'><tbody id='njDbst'></tbody></blockquote></table></option></ol><u id='njDbst'></u><kbd id='njDbst'><kbd id='njDbst'></kbd></kbd>

    <code id='njDbst'><strong id='njDbst'></strong></code>

    <fieldset id='njDbst'></fieldset>
          <span id='njDbst'></span>

              <ins id='njDbst'></ins>
              <acronym id='njDbst'><em id='njDbst'></em><td id='njDbst'><div id='njDbst'></div></td></acronym><address id='njDbst'><big id='njDbst'><big id='njDbst'></big><legend id='njDbst'></legend></big></address>

              <i id='njDbst'><div id='njDbst'><ins id='njDbst'></ins></div></i>
              <i id='njDbst'></i>
            1. <dl id='njDbst'></dl>
              1. <blockquote id='njDbst'><q id='njDbst'><noscript id='njDbst'></noscript><dt id='njDbst'></dt></q></blockquote><noframes id='njDbst'><i id='njDbst'></i>

                步態識別

                VersatileGait: A Large-Scale Synthetic Gait Dataset with Fine-Grained Attributes and Complicated Scenarios

                Submitted by neurta on Sun, 01/10/2021 - 16:25
                With the motivation of practical gait recognition applications, we propose to automatically create a large-scale synthetic gait dataset (called VersatileGait) by a game engine, which consists of around one million silhouette sequences of 11,000 subjects with fine-grained attributes in various complicated scenarios. Compared with existing real gait datasets with limited samples and simple scenarios, the proposed VersatileGait dataset possesses several nice properties, including huge dataset size, high sample diversity, high-quality annotations, multi-pitch angles, small domain gap with the real one, etc. Furthermore, we investigate the effectiveness of our dataset (e.g., domain transfer after pretraining). Then, we use the fine-grained attributes from VersatileGait to promote gait recognition in both accuracy and speed, and meanwhile justify the gait recognition performance under multi-pitch angle settings. Additionally, we explore a variety of potential applications for research. Extensive experiments demonstrate the value and effectiveness of the proposed VersatileGait in gait recognition along with its associated applications. We will release both VersatileGait and its corresponding data generation toolkit for further studies.