{"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/twinlitenetplus-a-stronger-model-for-real","title":"TwinLiteNetPlus: A Stronger Model for Real-time Drivable Area and Lane Segmentation","arxiv_id":"2403.16958","date":"2024-03-25","proceeding":null,"authors":["Quang-Huy Che","Duc-Tri Le","Minh-Quan Pham","Vinh-Tiep Nguyen","Duc-Khai Lam"],"abstract":"Semantic segmentation is crucial for autonomous driving, particularly for Drivable Area and Lane Segmentation, ensuring safety and navigation. To address the high computational costs of current state-of-the-art (SOTA) models, this paper introduces TwinLiteNetPlus (TwinLiteNet$^+$), a model adept at balancing efficiency and accuracy. TwinLiteNet$^+$ incorporates standard and depth-wise separable dilated convolutions, reducing complexity while maintaining high accuracy. It is available in four configurations, from the robust 1.94 million-parameter TwinLiteNet$^+_{\\text{Large}}$ to the ultra-compact 34K-parameter TwinLiteNet$^+_{\\text{Nano}}$. Notably, TwinLiteNet$^+_{\\text{Large}}$ attains a 92.9\\% mIoU for Drivable Area Segmentation and a 34.2\\% IoU for Lane Segmentation. These results notably outperform those of current SOTA models while requiring a computational cost that is approximately 11 times lower in terms of Floating Point Operations (FLOPs) compared to the existing SOTA model. Extensively tested on various embedded devices, TwinLiteNet$^+$ demonstrates promising latency and power efficiency, underscoring its suitability for real-world autonomous vehicle applications.","url_abs":"https://arxiv.org/abs/2403.16958v1","url_pdf":"https://arxiv.org/pdf/2403.16958v1.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":"twinlitenetplus-a-stronger-model-for-real","repo_url":"https://github.com/chequanghuy/TwinLiteNetPlus","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"twinlitenetplus-a-stronger-model-for-real","repo_url":"https://github.com/chequanghuy/TwinLiteNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"autonomous-driving","task_name":"Autonomous Driving"},{"task_slug":"drivable-area-detection","task_name":"Drivable Area Detection"},{"task_slug":"lane-detection","task_name":"Lane Detection"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/drivable-area-detection-on-bdd100k-val","task":"Drivable Area Detection","dataset":"BDD100K val","model":"TwinLiteNetPlus-Large","rank_in_archive_order":2,"of":10,"metrics":{"Params (M)":"1.94","mIoU":"92.9"},"uses_additional_data":false},{"leaderboard":"/sota/drivable-area-detection-on-bdd100k-val","task":"Drivable Area Detection","dataset":"BDD100K val","model":"TwinLiteNetPlus-Medium","rank_in_archive_order":4,"of":10,"metrics":{"Params (M)":"0.48","mIoU":"92.0"},"uses_additional_data":false},{"leaderboard":"/sota/drivable-area-detection-on-bdd100k-val","task":"Drivable Area Detection","dataset":"BDD100K val","model":"TwinLiteNetPlus-Small","rank_in_archive_order":8,"of":10,"metrics":{"Params (M)":"0.12","mIoU":"90.6"},"uses_additional_data":false},{"leaderboard":"/sota/drivable-area-detection-on-bdd100k-val","task":"Drivable Area Detection","dataset":"BDD100K val","model":"TwinLiteNetPlus-Nano","rank_in_archive_order":10,"of":10,"metrics":{"Params (M)":"0.03","mIoU":"87.3"},"uses_additional_data":false},{"leaderboard":"/sota/lane-detection-on-bdd100k-val","task":"Lane Detection","dataset":"BDD100K val","model":"TwinLiteNetPlus-Large","rank_in_archive_order":1,"of":11,"metrics":{"Accuracy (%)":"81.9","IoU (%)":"34.2","Params (M)":"1.94"},"uses_additional_data":false},{"leaderboard":"/sota/lane-detection-on-bdd100k-val","task":"Lane Detection","dataset":"BDD100K val","model":"TwinLiteNetPlus-Medium","rank_in_archive_order":2,"of":11,"metrics":{"Accuracy (%)":"79.1","IoU (%)":"32.3","Params (M)":"0.48"},"uses_additional_data":false},{"leaderboard":"/sota/lane-detection-on-bdd100k-val","task":"Lane Detection","dataset":"BDD100K val","model":"TwinLiteNetPlus-Small","rank_in_archive_order":6,"of":11,"metrics":{"Accuracy (%)":"75.8","IoU (%)":"29.3","Params (M)":"0.12"},"uses_additional_data":false},{"leaderboard":"/sota/lane-detection-on-bdd100k-val","task":"Lane Detection","dataset":"BDD100K val","model":"TwinLiteNetPlus-Nano","rank_in_archive_order":10,"of":11,"metrics":{"Accuracy (%)":"70.2","IoU (%)":"23.3","Params (M)":"0.03"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}