AI Paper Atlas

Research map / Video / Action recognition

Frame2Freq: Spectral Adapters for Fine-Grained Video Understanding

Thinesh Thiyakesan Ponbagavathi, Constantin Seibold, Alina Roitberg

CVPR 2026 · Video › Action recognition · approach: Fully supervised temporal action detection · 0 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 frequency conv. to temporal conv. branch (ablation)Frame2Freq-MS · Diving48top-1 accuracy: 90.4 -> 92.2Improved
add temporal conv. to frequency-only branch (ablation)Frame2Freq-MS · Diving48top-1 accuracy: 90.9 -> 92.2Improved
add temporal conv. to frequency-only (ablation)Frame2Freq-ST · Drive&Actaccuracy: 81.6 -> 82.0Improved
add frequency conv. to temporal-only (ablation)Frame2Freq-ST · Drive&Actaccuracy: 75.2 -> 82.0Improved
place adapter after MHSA instead of before (ablation)Frame2Freq · IKEA-ASMaccuracy: 76.1 -> 78.1Improved
place adapter before+after MHSA vs after only (ablation)Frame2Freq-MS · SSv2top-1 accuracy: 68.3 -> 69.7Improved
mean/concat fusion instead of gated fusion (ablation)Frame2Freq-MS · Diving48top-1 accuracy: 91.7 -> 92.2Improved
use three window scales instead of one (ablation)Frame2Freq-MS · Diving48top-1 accuracy: 91.5 -> 92.2Improved
add fourth window scale T/8 (ablation)Frame2Freq-MS · Diving48top-1 accuracy: 92.2 -> 91.0Hurt
place adapters in all depth ranges vs only layers 1-4 (ablation)Frame2Freq-MS · Diving48top-1 accuracy: 67.6 -> 92.2Improved
multi-scale variant instead of single-scale ST (ablation)Frame2Freq · Diving48top-1 accuracy: 75.1 -> 92.2Improved
replace ST-Adapter by Frame2Freq-MSST-Adapter · SSv2top-1 accuracy: 69.5 -> 70.4Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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