HOW DO THE DESIGN CHOICES, DATASET CONSTRUCTIONS, AND EVALUATION PRACTICES USED IN FACIAL EMOTION RECOGNITION RESEARCH LIMIT THE RELIABILITY OF THESE SYSTEMS OUTSIDE CONTROLLED BENCHMARK SETTINGS (2025)

Status: pending Scraped: 2026-08-12 Source: semantic_scholar Relevance: Background

Authors: Tanush Agrawal

Abstract

Facial emotion recognition systems are commonly evaluated using benchmark datasets and standard classification metrics. High accuracy on these benchmarks is often taken as evidence of reliability. This paper reviews how design choices,dataset construction, and evaluation practices shape that percept…

Key Contribution

Review and expand.

Design

Review and expand.

Key Results

Review and expand.

Impact

Review and expand.

Linked To

Add cross-references after review.