AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Investigating Self-Supervised Representations for Audio-Visual Deepfake Detection

Dragos-Alexandru Boldisor, Stefan Smeu, Dan Oneata, Elisabeta Oneata

CVPR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Self-supervised audio-visual representation learning · 1 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
use pretrained instead of random-init AV-HuBERT audio and video encoders (ablation)AV-HuBERT (A+V) synchronization · AV1MAUC (%): 50.0 -> 87.3Improved
use pretrained instead of random-init features for next-token prediction (ablation)AV-HuBERT (A+V) · AV1MAUC (%): 64.4 -> 84.5Improved
pair Wav2Vec2 with AV-HuBERT (V) rather than CLIP for synchronizationWav2Vec2 sync · AV1MAUC (%): 49.7 -> 86.5Improved
fuse CLIP with AV-HuBERT (A) for next-token predictionAV-HuBERT (A) NTP · AV1MAUC (%): 90.6 -> 86.9Hurt
use transformer head instead of linear head (ablation)FSFM features · FAVC (train AV1M)AUC (%): 40.9 -> 72.3Improved
use transformer head instead of linear head (ablation)AV-HuBERT (A) features · AV1M (train FAVC)AUC (%): 99.0 -> 88.8Hurt
use max pooling instead of average pooling (ablation)AV-HuBERT alignment, shift 0 · AV1MAUC (%): 65.3 -> 76.8Improved
use min pooling instead of average pooling (ablation)AV-HuBERT alignment, shift 0 · DFE-2024AUC (%): 62.7 -> 66.3Improved
allow temporal shift of 15 frames instead of 0 (ablation)AV-HuBERT alignment, average pooling · DFE-2024AUC (%): 62.7 -> 75.6Improved
allow temporal shift of 15 frames instead of 0 (ablation)AV-HuBERT alignment, min pooling · AV1MAUC (%): 58.1 -> 56.1Hurt
use layer 9 instead of the last layer of AV-HuBERT (ablation)SpeechForensics alignment · AV1MAUC (%): 67.1 -> 68.6No clear effect
use BRAVEn (V) visual features instead of Wav2Vec2 audio featureslinear probe · Average over FAVC, AVLips, DFEAUC (%): 71.6 -> 90.5Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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