Research map / Video / Video segmentation and tracking
AJILE Movement Prediction: Multimodal Deep Learning for Natural Human Neural Recordings and Video
Nancy Wang, Ali Farhadi, Rajesh Rao, Bingni Brunton
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 |
|---|---|---|---|
| combine ECoG and video with late fusion | AJILE movement classifier · AJILE Pred-b | accuracy: 64.9 -> 67.6 | Improved |
| combine ECoG and video with late fusion | AJILE movement classifier · AJILE Pred | accuracy: 61.8 -> 69.2 | Improved |
| combine ECoG and video with late fusion | AJILE movement classifier · AJILE Det | accuracy: 67.9 -> 72.7 | Improved |
| replace late fusion with early fusion (ablation) | AJILE multimodal model · AJILE Pred-b | accuracy: 67.6 -> 60.9 | Hurt |
| replace late fusion with early fusion (ablation) | AJILE multimodal model · AJILE Pred | accuracy: 69.2 -> 61.0 | Hurt |
| replace late fusion with naive averaging of modality outputs (ablation) | AJILE multimodal model · AJILE Pred | accuracy: 69.2 -> 66.8 | Hurt |
| add convolutional feature extraction before the LSTM (ablation) | AJILE ECoG model · AJILE Pred-b | accuracy: 49.1 -> 64.1 | Improved |
| add convolutional feature extraction before the LSTM (ablation) | AJILE ECoG model · AJILE Pred | accuracy: 51.9 -> 61.8 | Improved |
| replace 1D temporal filters with 3D spatiotemporal filters (ablation) | AJILE ECoG CNN · AJILE Pred-b | accuracy: 59.8 -> 55.7 | Hurt |
| replace 1D temporal filters with 3D spatiotemporal filters (ablation) | AJILE ECoG CNN · AJILE Pred | accuracy: 62.1 -> 57.3 | Hurt |
| use an ECoG deep model instead of an SVM with spectral features | AJILE ECoG movement classifier · AJILE Det | accuracy: 60.8 -> 70.3 | Improved |
| replace a single ECoG signal with its mean value (ablation) | AJILE multimodal model · AJILE single-electrode ablation tests | accuracy: ECoG only decreases: 2.6–26.9; multimodal decreases: 0.8–4.2 | Improved |
Closest papers by meaning
- ALIEN: Implicit Neural Representations for Human Motion Prediction under Arbitrary Latency (CVPR 2025 · similarity 0.85)
- HVIS: A Human-like Vision and Inference System for Human Motion Prediction (AAAI 2025 · similarity 0.85)
- emg2qwerty: A Large Dataset with Baselines for Touch Typing using Surface Electromyography (NeurIPS 2024 · similarity 0.84)
- Expressive Forecasting of 3D Whole-Body Human Motions (AAAI 2024 · similarity 0.84)
- EgoTac: In-the-wild Tactile Prediction from Egocentric Vision (NeurIPS 2026 · similarity 0.84)
- Forecasting of 3D Whole-body Human Poses with Grasping Objects (CVPR 2024 · similarity 0.84)
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.
