VideoJAM: Joint Appearance-Motion Representations for Enhanced Motion Generation in Video Models · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2502.02492/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTVideoJAM: Joint Appearance-Motion Representations for Enhanced Motion Generation in Video Models — mentioned by mdp2021https://news.ycombinator.com/item?id=43061729urn:hackerpapers:mention:2502.02492:43061729Sat, 15 Feb 2025 19:55:41 GMT<p>Website: https://hila-chefer.github.io/videojam-paper.github.io/ Paper: https://arxiv.org/abs/2502.02492 From this Two Minute Papers video about VideoJAM I understand this technology contains something akin to &quot;Attention is all you need&quot;, i.e. &quot;Inner Guidance is all you need&quot; (for motion in videos).</p><p><a href="https://hackerpaper.kaipfister.de/p/2502.02492/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.CV