{"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/learning-to-steer-by-mimicking-features-from","title":"Learning to Steer by Mimicking Features from Heterogeneous Auxiliary Networks","arxiv_id":"1811.02759","date":"2018-11-07","proceeding":null,"authors":["Yuenan Hou","Zheng Ma","Chunxiao Liu","Chen Change Loy"],"abstract":"The training of many existing end-to-end steering angle prediction models\nheavily relies on steering angles as the supervisory signal. Without learning\nfrom much richer contexts, these methods are susceptible to the presence of\nsharp road curves, challenging traffic conditions, strong shadows, and severe\nlighting changes. In this paper, we considerably improve the accuracy and\nrobustness of predictions through heterogeneous auxiliary networks feature\nmimicking, a new and effective training method that provides us with much\nricher contextual signals apart from steering direction. Specifically, we train\nour steering angle predictive model by distilling multi-layer knowledge from\nmultiple heterogeneous auxiliary networks that perform related but different\ntasks, e.g., image segmentation or optical flow estimation. As opposed to\nmulti-task learning, our method does not require expensive annotations of\nrelated tasks on the target set. This is made possible by applying contemporary\noff-the-shelf networks on the target set and mimicking their features in\ndifferent layers after transformation. The auxiliary networks are discarded\nafter training without affecting the runtime efficiency of our model. Our\napproach achieves a new state-of-the-art on Udacity and Comma.ai, outperforming\nthe previous best by a large margin of 12.8% and 52.1%, respectively.\nEncouraging results are also shown on Berkeley Deep Drive (BDD) dataset.","url_abs":"http://arxiv.org/abs/1811.02759v1","url_pdf":"http://arxiv.org/pdf/1811.02759v1.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":"learning-to-steer-by-mimicking-features-from","repo_url":"https://github.com/aj9011/Car-Speed-Prediction","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"learning-to-steer-by-mimicking-features-from","repo_url":"https://github.com/cardwing/Codes-for-Steering-Control","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"image-segmentation","task_name":"Image Segmentation"},{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"},{"task_slug":"optical-flow-estimation","task_name":"Optical Flow Estimation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"steering-control","task_name":"Steering Control"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/steering-control-on-bdd100k-val","task":"Steering Control","dataset":"BDD100K val","model":"FM-Net","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"85.03"},"uses_additional_data":false},{"leaderboard":"/sota/steering-control-on-commaai","task":"Steering Control","dataset":"Comma.ai","model":"FM-Net","rank_in_archive_order":1,"of":1,"metrics":{"MAE":"0.7048"},"uses_additional_data":false},{"leaderboard":"/sota/steering-control-on-udacity","task":"Steering Control","dataset":"Udacity","model":"FM-Net","rank_in_archive_order":1,"of":1,"metrics":{"MAE":"1.6236"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.02759","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1811.02759"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/aj9011/Car-Speed-Prediction","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/cardwing/Codes-for-Steering-Control","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":6},"by_repo_kind":{"listed":{"samples":6,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"83877d60072f32f1","entry":"conv3x3","repo":"cardwing/Codes-for-Steering-Control","repo_kind":"listed","path":"semantic-segmentation/models/fc_resnet.py","file_url":"https://github.com/cardwing/Codes-for-Steering-Control/blob/HEAD/semantic-segmentation/models/fc_resnet.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"83877d60072f32f1"}},{"code_sha256_prefix":"7e9da385f15a6336","entry":"fc_resnet101","repo":"cardwing/Codes-for-Steering-Control","repo_kind":"listed","path":"semantic-segmentation/models/fc_sense_resnet.py","file_url":"https://github.com/cardwing/Codes-for-Steering-Control/blob/HEAD/semantic-segmentation/models/fc_sense_resnet.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7e9da385f15a6336"}},{"code_sha256_prefix":"242966ab89247bec","entry":"fc_resnet18","repo":"cardwing/Codes-for-Steering-Control","repo_kind":"listed","path":"semantic-segmentation/models/fc_resnet.py","file_url":"https://github.com/cardwing/Codes-for-Steering-Control/blob/HEAD/semantic-segmentation/models/fc_resnet.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"242966ab89247bec"}},{"code_sha256_prefix":"c2a3f3279881cb3d","entry":"fc_resnet34","repo":"cardwing/Codes-for-Steering-Control","repo_kind":"listed","path":"semantic-segmentation/models/fc_resnet.py","file_url":"https://github.com/cardwing/Codes-for-Steering-Control/blob/HEAD/semantic-segmentation/models/fc_resnet.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c2a3f3279881cb3d"}},{"code_sha256_prefix":"cfd74bf2d4f2cef3","entry":"inference_small_config","repo":"cardwing/Codes-for-Steering-Control","repo_kind":"listed","path":"steering-control/resnet.py","file_url":"https://github.com/cardwing/Codes-for-Steering-Control/blob/HEAD/steering-control/resnet.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cfd74bf2d4f2cef3"}},{"code_sha256_prefix":"94defdbecd90aed8","entry":"loss","repo":"cardwing/Codes-for-Steering-Control","repo_kind":"listed","path":"steering-control/resnet.py","file_url":"https://github.com/cardwing/Codes-for-Steering-Control/blob/HEAD/steering-control/resnet.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"94defdbecd90aed8"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}