Advances in Causal Analysis and Visual Understanding

The field of affective computing and visual understanding is moving towards a greater emphasis on causal analysis and discovery. Researchers are developing new frameworks and models that can infer causal relationships between variables, such as facial expressions and emotions, and identify the underlying mechanisms that drive these relationships. This has led to significant advances in areas such as facial affect analysis, micro-expression intensity estimation, and visual causal discovery. Notable papers in this area include CausalAffect, which proposes a framework for causal graph discovery in facial affect analysis, and CauSight, which introduces a novel vision-language model for visual causal discovery. Additionally, papers such as Weakly Supervised Continuous Micro-Expression Intensity Estimation Using Temporal Deep Neural Network and Mitigating Gender Bias in Depression Detection via Counterfactual Inference have made significant contributions to the field by addressing important challenges such as data limitation and bias.

Sources

CausalAffect: Causal Discovery for Facial Affective Understanding

Weakly Supervised Continuous Micro-Expression Intensity Estimation Using Temporal Deep Neural Network

CauSight: Learning to Supersense for Visual Causal Discovery

Mitigating Gender Bias in Depression Detection via Counterfactual Inference

LightHCG: a Lightweight yet powerful HSIC Disentanglement based Causal Glaucoma Detection Model framework

EMINDS: Understanding User Behavior Progression for Mental Health Exploration on Social Media

Measuring the Unspoken: A Disentanglement Model and Benchmark for Psychological Analysis in the Wild

Built with on top of