{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/stepwise-goal-driven-networks-for-trajectory","title":"Stepwise Goal-Driven Networks for Trajectory Prediction","arxiv_id":"2103.14107","date":"2021-03-25","proceeding":null,"authors":["Chuhua Wang","Yuchen Wang","Mingze Xu","David J. Crandall"],"abstract":"We propose to predict the future trajectories of observed agents (e.g., pedestrians or vehicles) by estimating and using their goals at multiple time scales. We argue that the goal of a moving agent may change over time, and modeling goals continuously provides more accurate and detailed information for future trajectory estimation. To this end, we present a recurrent network for trajectory prediction, called Stepwise Goal-Driven Network (SGNet). Unlike prior work that models only a single, long-term goal, SGNet estimates and uses goals at multiple temporal scales. In particular, it incorporates an encoder that captures historical information, a stepwise goal estimator that predicts successive goals into the future, and a decoder that predicts future trajectory. We evaluate our model on three first-person traffic datasets (HEV-I, JAAD, and PIE) as well as on three bird's eye view datasets (NuScenes, ETH, and UCY), and show that our model achieves state-of-the-art results on all datasets. Code has been made available at: https://github.com/ChuhuaW/SGNet.pytorch.","url_abs":"https://arxiv.org/abs/2103.14107v3","url_pdf":"https://arxiv.org/pdf/2103.14107v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"stepwise-goal-driven-networks-for-trajectory","repo_url":"https://github.com/ChuhuaW/SGNet.pytorch","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"multi-future-trajectory-prediction","task_name":"Multi-future Trajectory Prediction"},{"task_slug":"prediction","task_name":"Prediction"},{"task_slug":"trajectory-prediction","task_name":"Trajectory Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/trajectory-prediction-on-ethucy","task":"Trajectory Prediction","dataset":"ETH/UCY","model":"SGNet","rank_in_archive_order":4,"of":20,"metrics":{"ADE-8/12":"0.18","FDE-8/12":"0.35"},"uses_additional_data":false},{"leaderboard":"/sota/trajectory-prediction-on-hev-i","task":"Trajectory Prediction","dataset":"HEV-I","model":"SGNet","rank_in_archive_order":1,"of":2,"metrics":{"ADE(0.5)":"6.28","ADE(1.0)":"11.35","ADE(1.5)":"18.27","FDE(1.5)":"39.86","FIOU(1.5)":"0.63"},"uses_additional_data":false},{"leaderboard":"/sota/trajectory-prediction-on-jaad","task":"Trajectory Prediction","dataset":"JAAD","model":"SGNet","rank_in_archive_order":1,"of":5,"metrics":{"CF_MSE(1.5)":"4076","C_MSE(1.5)":"996","MSE(0.5)":"82","MSE(1.0)":"328","MSE(1.5)":"1049"},"uses_additional_data":false},{"leaderboard":"/sota/trajectory-prediction-on-pie","task":"Trajectory Prediction","dataset":"PIE","model":"SGNet","rank_in_archive_order":1,"of":5,"metrics":{"CF_MSE(1.5)":"1761","C_MSE(1.5)":"413","MSE(0.5)":"34","MSE(1.0)":"133","MSE(1.5)":"442"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2103.14107","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}