Research map / Architectures and efficiency / Attention and transformers
EEGiT: Teaching Vision Transformers to Understand the EEG signal
Jiahao Zhou, Chenghao Xu, Wei Wang, Erkun Yang, Cheng Deng
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 |
|---|---|---|---|
| add both pretrained ViT weights and EEG patch representation (ablation) | ViT EEG encoder · THINGS-EEG intra-subject | top-1 accuracy: 41.8 -> 70.4 | Improved |
| add both pretrained ViT weights and EEG patch representation (ablation) | ViT EEG encoder · THINGS-EEG inter-subject | top-1 accuracy: 17.4 -> 24.0 | Improved |
| remove one of the two components (second ablation row) (ablation) | EEGiT · THINGS-EEG intra-subject | top-1 accuracy: 70.4 -> 63.6 | Hurt |
| remove the other component (third ablation row) (ablation) | EEGiT · THINGS-EEG intra-subject | top-1 accuracy: 70.4 -> 54.0 | Hurt |
| add both pretrained ViT and EEG patch (ablation) | ViT EEG encoder · EEG-3D color classification | top-1 accuracy: 34.26 -> 41.32 | Improved |
| add both pretrained ViT and EEG patch (ablation) | ViT EEG encoder · EEG-3D object classification | top-1 accuracy: 4.28 -> 5.95 | Improved |
| initialize with pretrained ViT weights rather than random (ablation) | EEGiT · THINGS-EEG retrieval | convergence speed: about 10 epochs vs about 20 epochs | Improved |
| keep only occipital channels instead of all channels (ablation) | EEGiT · THINGS-EEG retrieval | top-1/top-5 accuracy | No clear effect |
| remove occipital region signals (ablation) | EEGiT · THINGS-EEG retrieval | top-1/top-5 accuracy | Hurt |
| EEGiT versus Neuro-3D on object classification | EEGiT · EEG-3D object type | top-1 accuracy: 5.91 -> 5.95 | No clear effect |
Closest papers by meaning
- What Does the Brain See? Multiview Neural Representation to Demystify the Brain-Visual Alignment (NeurIPS 2026 · similarity 0.91)
- MindAlign: Bridging EEG, Vision, and Language for Zero-Shot Visual Decoding (NeurIPS 2026 · similarity 0.90)
- Decoding Natural Images from EEG for Object Recognition (ICLR 2024 · similarity 0.90)
- Visual Decoding and Reconstruction via EEG Embeddings with Guided Diffusion (NeurIPS 2024 · similarity 0.90)
- Neuro-3D: Towards 3D Visual Decoding from EEG Signals (CVPR 2025 · similarity 0.89)
- EEG2Video: Towards Decoding Dynamic Visual Perception from EEG Signals (NeurIPS 2024 · similarity 0.89)
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.
