AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

[Re] Reproducibility Report: Contrastive Learning of Socially-aware Motion Representations

Roopsa Sen, Sidharth Sinha, Animesh Jha, Parv Maheshwari

NeurIPS 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
tune Social-NCE hyperparameters (tau, horizon, lambda, augmentation) (ablation)Social-STGCNN · ETHcollision rate (COL): 3.54 -> 3.45Improved
tune Social-NCE hyperparameters (ablation)Social-STGCNN · ETHFDE: 0.678 -> 0.674Improved
raise contrastive weight lambda from 2 to 16 (ablation)Social-STGCNN · ETHtuned loss: 2 -> 16Improved
add Social-NCE lossTrajectron++ · ETH/UCY averagecollision rate: 35.7% decreaseImproved
add Social-NCE lossSocial-STGCNN · ETH/UCY averagecollision rate: 35.1% decreaseImproved
add Social-NCE lossimitation learning · robot navigationcollision rate: 68.9% decreaseImproved
add Social-NCE losstrajectory models · ETH/UCYFDE: less than 1% deviationNo clear effect
add Social-NCE loss to Rainbow DQNreinforcement learning · simulated crowd navigationepisodes to reward 0.6: 4000 -> 2000 episodesImproved
port code to PyTorch LightningSocial-STGCNN · training timetraining time: 8hr 32mins -> 8hr 30minsNo clear effect
port code to PyTorch LightningTrajectron++ · training timetraining time: 4hr 28mins -> 4hr 24minsNo clear effect
port code to PyTorch LightningSocial-STGCNN · ETHcollision rate: 1.11 -> 0.53Mixed
vary data augmentation hyperparameters (ablation)Social-NCE · ETH/UCYaccuracyMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy