AI Paper Atlas

Research map / Benchmarks / RAF-DB

RAF-DB leaderboard

16 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on RAF-DB. Best accuracy: 94.04 (AURA (Ours), ICML 2026).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

RAF-DB · accuracy

16 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
194.04AURA (Ours)AURA: Visually Interpretable Affective Understanding via Robust Archetypes
ICML 2026 · Table 1
Human and embodied modeling
293.74MPA-FER
CLIP ViT-L/14
Multimodal Prompt Alignment for Facial Expression Recognition
ICCV 2025 · Table 6
CLIP and zero-shot
393.02QCS (pretrained on AffectNet-8)
IR50
QCS:Feature Refining from Quadruplet Cross Similarity for Facial Expression Recognition
AAAI 2025 · Comparison with the State of t
Recognition and long-tail
492.97NorfaceNorface: Improving Facial Expression Analysis by Identity Normalization
ECCV 2024 · Table 4
Recognition and long-tail
592.05EmoLA
LLaVA-1.5 7B
Facial Affective Behavior Analysis with Instruction Tuning
ECCV 2024 · Table 1
Multimodal LLMs
691.53ARM + Ours
ResNet-18
Attack Can Benefit: An Adversarial Approach to Recognizing Facial Expressions under Noisy Annotations
AAAI 2023 · Table 1
Attacks on text and graphs
790.91TransFER
IR-50 + ViT
TransFER: Learning Relation-Aware Facial Expression Representations With Transformers
ICCV 2021 · Table 7
Attention and transformers
890.33Ours (concepPath)
ResNet-50
A Brain-Inspired Way of Reducing the Network Complexity via Concept-Regularized Coding for Emotion Recognition
AAAI 2024 · Table 1
Human and embodied modeling
990.09CLEF
FaRL ViT-B/16
Weakly-Supervised Text-Driven Contrastive Learning for Facial Behavior Understanding
ICCV 2023 · Table 5
Recognition and long-tail
1089.95FRA (FT)
ResNet-50
Self-Supervised Facial Representation Learning with Facial Region Awareness
CVPR 2024 · Table 2
Self-supervised and contrastive
6 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.