AI Paper Atlas

Research map / Video / Video segmentation and tracking

Cinematic Mindscapes: High-quality Video Reconstruction from Brain Activity

Zijiao Chen, Jiaxin Qing, Juan Helen Zhou

NeurIPS 2023 · Video › Video segmentation and tracking · approach: Sign language and physiological video analysis · 20 citations (OpenAlex)

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
add multimodal (text+image) contrastive learning (ablation)Mind-Video · Wen (2018) Subject 1, frame-based 50-wayaccuracy: 0.088 -> 0.172Improved
add multimodal (text+image) contrastive learning (ablation)Mind-Video · Wen (2018) Subject 1, SSIMSSIM: 0.135 -> 0.171Improved
use both text and image contrastive targets instead of text only (ablation)Mind-Video · Wen (2018) Subject 1, video-based 50-wayaccuracy: 0.185 -> 0.202Improved
use both text and image contrastive targets instead of image only (ablation)Mind-Video · Wen (2018) Subject 1, video-based 50-wayaccuracy: 0.189 -> 0.202Improved
use spatiotemporal window size 2 instead of 1 (ablation)Mind-Video · Wen (2018) Subject 1, frame-based 2-wayaccuracy: 0.759 -> 0.792Improved
use spatiotemporal window size 2 instead of 1 (ablation)Mind-Video · Wen (2018) Subject 1, SSIMSSIM: 0.169 -> 0.171Improved
use window size 2 instead of 3 (ablation)Mind-Video · Wen (2018) Subject 1, video-based 50-wayaccuracy: 0.161 -> 0.202Improved
add adversarial guidance at inference (ablation)Mind-Video · Wen (2018) Subject 1, frame-based 50-wayaccuracy: 0.117 -> 0.172Improved
add adversarial guidance at inference (ablation)Mind-Video · Wen (2018) Subject 1, SSIMSSIM: 0.152 -> 0.171Improved
add adversarial guidance at inference (ablation)Mind-Video · Wen (2018) Subject 1, video-based 2-wayaccuracy: 0.859 -> 0.853No clear effect

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