Tag: #I3D
Kubernetes, DevOps and MLOps, written from things we ran and measured ourselves · 1 posts
Video Models: 3D Convolutions, Two-Stream (SlowFast), Attention (TimeSformer) — and Masked Video Modeling (VideoMAE)
We break down three ways to add time to a single-image model: 3D convolutions (C3D, I3D starting from inflated ImageNet weights), two-stream (SlowFast) running slow and fast paths side by side, and transformers with fact
2026-10-17 · 16 min read #비전 모델#영상 인식#I3D#SlowFast#TimeSformer