Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386245243> ?p ?o ?g. }
- W4386245243 abstract "Current LiDAR-based 3D object detectors for autonomous driving are almost entirely trained on human-annotated data collected in specific geographical domains with specific sensor setups, making it difficult to adapt to a different domain. MODEST [1] is the first work to train 3D object detectors without any labels. Our work, HyperMODEST, proposes a universal method implemented on top of MODEST that can largely accelerate the self-training process and does not require tuning on a specific dataset. We filter intermediate pseudo-labels used for data augmentation with low confidence scores. On the nuScenes dataset, we observe a significant improvement of 1.6% in AP_BEV in 0-80m range at IoU=0.25 and an improvement of 1.7% in AP BEV in 0-80m range at IoU=0.5 while only using one-fifth of the training time in the original approach by MODEST [1]. On the Lyft dataset, we also observe an improvement over the baseline during the first round of iterative self-training. We explore the trade-off between high precision and high recall in the early stage of the self-training process by comparing our proposed method with two other score filtering methods: confidence score filtering for pseudo-labels with and without static label retention. The code and models of this work are available at https://github.com/TRAILab/HyperMODEST" @default.
- W4386245243 created "2023-08-30" @default.
- W4386245243 creator A5024242059 @default.
- W4386245243 creator A5084138424 @default.
- W4386245243 date "2023-06-01" @default.
- W4386245243 modified "2023-09-27" @default.
- W4386245243 title "HyperMODEST: Self-Supervised 3D Object Detection with Confidence Score Filtering" @default.
- W4386245243 cites W1521527082 @default.
- W4386245243 cites W1567556022 @default.
- W4386245243 cites W203234824 @default.
- W4386245243 cites W2150066425 @default.
- W4386245243 cites W2885139206 @default.
- W4386245243 cites W2895281799 @default.
- W4386245243 cites W2897529137 @default.
- W4386245243 cites W2949708697 @default.
- W4386245243 cites W2963400571 @default.
- W4386245243 cites W2985406498 @default.
- W4386245243 cites W3034975685 @default.
- W4386245243 cites W3035574168 @default.
- W4386245243 cites W3035750252 @default.
- W4386245243 cites W3106895564 @default.
- W4386245243 cites W3108884723 @default.
- W4386245243 cites W3126960890 @default.
- W4386245243 cites W3167095230 @default.
- W4386245243 cites W3175269419 @default.
- W4386245243 cites W3183392001 @default.
- W4386245243 cites W4214526701 @default.
- W4386245243 cites W4285102119 @default.
- W4386245243 cites W4312253363 @default.
- W4386245243 cites W4313123540 @default.
- W4386245243 cites W4317548422 @default.
- W4386245243 cites W4386249619 @default.
- W4386245243 doi "https://doi.org/10.1109/crv60082.2023.00035" @default.
- W4386245243 hasPublicationYear "2023" @default.
- W4386245243 type Work @default.
- W4386245243 citedByCount "0" @default.
- W4386245243 crossrefType "proceedings-article" @default.
- W4386245243 hasAuthorship W4386245243A5024242059 @default.
- W4386245243 hasAuthorship W4386245243A5084138424 @default.
- W4386245243 hasConcept C100660578 @default.
- W4386245243 hasConcept C106131492 @default.
- W4386245243 hasConcept C111368507 @default.
- W4386245243 hasConcept C111919701 @default.
- W4386245243 hasConcept C119857082 @default.
- W4386245243 hasConcept C12725497 @default.
- W4386245243 hasConcept C127313418 @default.
- W4386245243 hasConcept C138885662 @default.
- W4386245243 hasConcept C153180895 @default.
- W4386245243 hasConcept C154945302 @default.
- W4386245243 hasConcept C159985019 @default.
- W4386245243 hasConcept C177264268 @default.
- W4386245243 hasConcept C192562407 @default.
- W4386245243 hasConcept C199360897 @default.
- W4386245243 hasConcept C204323151 @default.
- W4386245243 hasConcept C2776151529 @default.
- W4386245243 hasConcept C2776760102 @default.
- W4386245243 hasConcept C2781238097 @default.
- W4386245243 hasConcept C31972630 @default.
- W4386245243 hasConcept C41008148 @default.
- W4386245243 hasConcept C41895202 @default.
- W4386245243 hasConcept C76155785 @default.
- W4386245243 hasConcept C81669768 @default.
- W4386245243 hasConcept C94915269 @default.
- W4386245243 hasConcept C98045186 @default.
- W4386245243 hasConceptScore W4386245243C100660578 @default.
- W4386245243 hasConceptScore W4386245243C106131492 @default.
- W4386245243 hasConceptScore W4386245243C111368507 @default.
- W4386245243 hasConceptScore W4386245243C111919701 @default.
- W4386245243 hasConceptScore W4386245243C119857082 @default.
- W4386245243 hasConceptScore W4386245243C12725497 @default.
- W4386245243 hasConceptScore W4386245243C127313418 @default.
- W4386245243 hasConceptScore W4386245243C138885662 @default.
- W4386245243 hasConceptScore W4386245243C153180895 @default.
- W4386245243 hasConceptScore W4386245243C154945302 @default.
- W4386245243 hasConceptScore W4386245243C159985019 @default.
- W4386245243 hasConceptScore W4386245243C177264268 @default.
- W4386245243 hasConceptScore W4386245243C192562407 @default.
- W4386245243 hasConceptScore W4386245243C199360897 @default.
- W4386245243 hasConceptScore W4386245243C204323151 @default.
- W4386245243 hasConceptScore W4386245243C2776151529 @default.
- W4386245243 hasConceptScore W4386245243C2776760102 @default.
- W4386245243 hasConceptScore W4386245243C2781238097 @default.
- W4386245243 hasConceptScore W4386245243C31972630 @default.
- W4386245243 hasConceptScore W4386245243C41008148 @default.
- W4386245243 hasConceptScore W4386245243C41895202 @default.
- W4386245243 hasConceptScore W4386245243C76155785 @default.
- W4386245243 hasConceptScore W4386245243C81669768 @default.
- W4386245243 hasConceptScore W4386245243C94915269 @default.
- W4386245243 hasConceptScore W4386245243C98045186 @default.
- W4386245243 hasLocation W43862452431 @default.
- W4386245243 hasOpenAccess W4386245243 @default.
- W4386245243 hasPrimaryLocation W43862452431 @default.
- W4386245243 hasRelatedWork W1988485990 @default.
- W4386245243 hasRelatedWork W2095705906 @default.
- W4386245243 hasRelatedWork W2127769904 @default.
- W4386245243 hasRelatedWork W2791133342 @default.
- W4386245243 hasRelatedWork W2801801420 @default.
- W4386245243 hasRelatedWork W2970686063 @default.
- W4386245243 hasRelatedWork W2975200075 @default.
- W4386245243 hasRelatedWork W2998858795 @default.