Research map / 3D vision / 3D generation
Egocentric Prediction of Action Target in 3D
Yiming Li, Ziang Cao, Andrew Liang, Benjamin Liang, Luoyao Chen, Hang Zhao, Chen Feng
Check 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 changed | Where | Measured | Effect |
|---|---|---|---|
| add TF and IMU features to visual features (ablation) | EgoPAT3D LSTM predictor · EgoPAT3D seen scenes | overall CLE (cm): 19.15 -> 18.61 | Improved |
| add TF and IMU features to visual features (ablation) | EgoPAT3D LSTM predictor · EgoPAT3D unseen scenes | overall CLE (cm): 19.61 -> 18.82 | Improved |
| omit TF motion features (ablation) | EgoPAT3D LSTM predictor · EgoPAT3D seen scenes | overall CLE (cm): 18.61 -> 21.81 | Hurt |
| omit IMU features (ablation) | EgoPAT3D LSTM predictor · EgoPAT3D seen scenes | overall CLE (cm): 18.61 -> 19.48 | Hurt |
| use TF+IMU without visual features (ablation) | EgoPAT3D predictor · EgoPAT3D seen scenes | overall CLE (cm): 19.15 -> 18.66 | Improved |
| apply stronger loss weighting early rather than late (ablation) | EgoPAT3D predictor · EgoPAT3D seen scenes | overall CLE (cm): 18.97 -> 18.61 | Improved |
| replace NLLLoss with TWRLoss (ablation) | EgoPAT3D predictor · EgoPAT3D seen scenes | overall CLE (cm): 21.63 -> 18.61 | Improved |
| replace NLLLoss with TWRLoss (ablation) | EgoPAT3D predictor · EgoPAT3D unseen scenes | overall CLE (cm): 21.71 -> 18.82 | Improved |
| replace GRU with LSTM (ablation) | EgoPAT3D predictor · EgoPAT3D seen scenes | overall CLE (cm): 20.07 -> 18.61 | Improved |
| replace GRU with LSTM (ablation) | EgoPAT3D predictor · EgoPAT3D unseen scenes | overall CLE (cm): 19.77 -> 18.82 | Improved |
| increase grid granularity from 1024^3/m^3 to 3072^3/m^3 (ablation) | EgoPAT3D predictor · EgoPAT3D seen scenes | overall CLE (cm): 19.88 -> 19.04 | Improved |
| increase grid granularity from 3072^3/m^3 to 5120^3/m^3 (ablation) | EgoPAT3D predictor · EgoPAT3D seen scenes | overall CLE (cm): 19.04 -> 18.61 | Improved |
Closest papers by meaning
- Generating 6DoF Object Manipulation Trajectories from Action Description in Egocentric Vision (CVPR 2025 · similarity 0.89)
- Multi-Dataset, Multitask Learning of Egocentric Vision Tasks (TPAMI 2021 · similarity 0.88)
- Egocentric Activity Recognition and Localization on a 3D Map (ECCV 2022 · similarity 0.88)
- EgoM2P: Egocentric Multimodal Multitask Pretraining (ICCV 2025 · similarity 0.88)
- Interactive Prototype Learning for Egocentric Action Recognition (ICCV 2021 · similarity 0.88)
- Gaze Beyond the Frame: Forecasting Egocentric 3D Visual Span (NeurIPS 2025 · similarity 0.87)
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.
