Research map / Probabilistic and scientific modeling / Causal discovery

CausalAffect: Causally Guided Learning of Psychology-Aligned Facial Affect Relations

NeurIPS 2026 · Probabilistic and scientific modeling › Causal discovery · approach: Causal representation identifiability

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes CausalAffect, a causally guided learning framework that models facial Action Unit relations (facilitation/inhibition) and links AU configurations to expression states, grounded in psychology theories.

Problem

Facial affect recognition is usually treated as image-to-label prediction, ignoring the causal relations among activated Action Units and how AU configurations produce expression states.

Approach

Treats facial affect understanding as modeling causal relations between AUs (facilitation and inhibition) and between AU configurations and expression states, guided by psychological theories (FACS, Component Process Model, motor synergy theories).

Experimental setup

Builds on: FACS Component Process Model

Closest papers by meaning

• Facial-R1: Aligning Reasoning and Recognition for Facial Emotion Analysis (AAAI 2026 · similarity 0.89)
• AFFAKT: A Hierarchical Optimal Transport Based Method for Affective Facial Knowledge Transfer in Video Deception Detection (AAAI 2025 · similarity 0.86)
• EmoNet-Face: An Expert-Annotated Benchmark for Synthetic Emotion Recognition (NeurIPS 2025 · similarity 0.86)
• VidEmo: Affective-Tree Reasoning for Emotion-Centric Video Foundation Models (NeurIPS 2025 · similarity 0.86)
• Robust Emotion Recognition in Context Debiasing (CVPR 2024 · similarity 0.85)
• A Brain-Inspired Way of Reducing the Network Complexity via Concept-Regularized Coding for Emotion Recognition (AAAI 2024 · similarity 0.85)