{"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/sparse-representations-for-object-and-ego","title":"Sparse Representations for Object and Ego-motion Estimation in Dynamic Scenes","arxiv_id":"1903.03731","date":"2019-03-09","proceeding":null,"authors":["Hirak J. Kashyap","Charless Fowlkes","Jeffrey L. Krichmar"],"abstract":"Dynamic scenes that contain both object motion and egomotion are a challenge\nfor monocular visual odometry (VO). Another issue with monocular VO is the\nscale ambiguity, i.e. these methods cannot estimate scene depth and camera\nmotion in real scale. Here, we propose a learning based approach to predict\ncamera motion parameters directly from optic flow, by marginalizing depthmap\nvariations and outliers. This is achieved by learning a sparse overcomplete\nbasis set of egomotion in an autoencoder network, which is able to eliminate\nirrelevant components of optic flow for the task of camera parameter or\nmotionfield estimation. The model is trained using a sparsity regularizer and a\nsupervised egomotion loss, and achieves the state-of-the-art performances on\ntrajectory prediction and camera rotation prediction tasks on KITTI and Virtual\nKITTI datasets, respectively. The sparse latent space egomotion representation\nlearned by the model is robust and requires only 5% of the hidden layer neurons\nto maintain the best trajectory prediction accuracy on KITTI dataset.\nAdditionally, in presence of depth information, the proposed method\ndemonstrates faithful object velocity prediction for wide range of object sizes\nand speeds by global compensation of predicted egomotion and a divisive\nnormalization procedure.","url_abs":"http://arxiv.org/abs/1903.03731v1","url_pdf":"http://arxiv.org/pdf/1903.03731v1.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":"sparse-representations-for-object-and-ego","repo_url":"https://github.com/hkashyap/SparseMotion","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"depth-and-camera-motion","task_name":"Depth And Camera Motion"},{"task_slug":"monocular-visual-odometry","task_name":"Monocular Visual Odometry"},{"task_slug":"motion-estimation","task_name":"Motion Estimation"},{"task_slug":"object","task_name":"Object"},{"task_slug":"prediction","task_name":"Prediction"},{"task_slug":"trajectory-prediction","task_name":"Trajectory Prediction"},{"task_slug":"visual-odometry","task_name":"Visual Odometry"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}