Research map / 3D vision / 3D generation
Haptic Neural Fields: Bringing Tactile Interactions to 3D Rendered Scenes
Antonio Luigi Stefani, Niccolò Bisagno, Nicola Conci, Eckehard Steinbach, Francesco De Natale
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 changed | Where | Measured | Effect |
|---|---|---|---|
| train contrastive encoders on mixed GelSight and DIGIT domains instead of GACM haptic baseline | haptic embedding with linear probe · HaTT material classification | accuracy (%): 77.29 -> 92.13 | Improved |
| mixed-domain contrastive training versus TaRF baseline for vision embeddings | vision embedding with linear probe · TnG material classification | accuracy (%): 57.60 -> 76.16 | Improved |
| balance TnG training data to 500 samples per class (ablation) | vision embeddings · HaTT material classification | accuracy (%): 68.75 -> 97.19 | Improved |
| balance TnG training data to 500 samples per class (ablation) | vision embeddings · TnG material classification | accuracy (%): 76.16 -> 75.47 | No clear effect |
| balance TnG training data to 500 samples per class (ablation) | haptic embeddings · TnG material classification | accuracy (%): 55.40 -> 67.73 | Improved |
| replace GACM with HNF action-conditioned neural field | vibrotactile synthesis · GelSight (HaTT) | MSE: 6557.32 -> 3443.15 | Improved |
| replace GACM with HNF | vibrotactile synthesis · GelSight (HaTT) | ST-SIM: 0.85 -> 0.85 | No clear effect |
| replace GACM with HNF | vibrotactile synthesis · GelSight (HaTT) | LSD: 0.80 -> 0.88 | Hurt |
| add action augmentation with eight synthetic primitives (HNF+) (ablation) | HNF · DIGIT | ST-SIM: 0.86 -> 0.88 | Improved |
| add action augmentation with eight synthetic primitives (HNF+) (ablation) | HNF · DIGIT | MSE: 4166.64 -> 3764.00 | Improved |
| add action augmentation with eight synthetic primitives (HNF+) (ablation) | HNF · DIGIT | LSD: 0.89 -> 0.89 | No clear effect |
Closest papers by meaning
- Tactile-Augmented Radiance Fields (CVPR 2024 · similarity 0.89)
- PhysTacGen: Physics-Aware Visual-Tactile Sensor Image Generation (NeurIPS 2026 · similarity 0.88)
- Tactile DreamFusion: Exploiting Tactile Sensing for 3D Generation (NeurIPS 2024 · similarity 0.88)
- Universal Visuo-Tactile Video Understanding for Embodied Interaction (NeurIPS 2025 · similarity 0.87)
- Touch2Shape: Touch-Conditioned 3D Diffusion for Shape Exploration and Reconstruction (CVPR 2025 · similarity 0.86)
- AnyTouch: Learning Unified Static-Dynamic Representation across Multiple Visuo-tactile Sensors (ICLR 2025 · similarity 0.86)
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.
