FED-Bench: A Cross-Granular Benchmark for Disentangled Evaluation of Facial Expression Editing

2026-03-31

AI Summary

Existing facial expression editing benchmarks are hindered by low-quality images, ambiguous instructions, and biased evaluation metrics, making it difficult to simultaneously preserve identity and achieve precise expression control. To address these limitations, this work introduces a high-quality benchmark dataset comprising 747 triplets and proposes FED-Score, a cross-granularity evaluation protocol that decouples performance into three dimensions: instruction alignment, fidelity, and expression gain. Leveraging a cascaded, scalable data construction pipeline and a large-scale in-the-wild training set, we conduct a systematic evaluation of 18 models, uncovering their bottlenecks in fine-grained instruction following. Further fine-tuning of baseline models substantially improves editing performance, demonstrating the effectiveness and extensibility of the proposed benchmark.

Abstract

Facial expression image editing requires fine-grained control to strictly preserve human identity and background while precisely manipulating expression. However, existing editing benchmarks primarily focus on general scenarios, lacking high-quality facial images and corresponding editing instructions. Furthermore, current evaluation metrics exhibit systemic biases in this task, often favoring lazy editing or overfit editing. To bridge these gaps, we propose FED-Bench, a comprehensive benchmark featuring rigorous testing and an accurate evaluation suite. First, we carefully construct a benchmark of 747 triplets through a cascaded and scalable pipeline, each comprising an original image, an editing instruction, and a ground-truth image for precise evaluation. Second, we introduce FED-Score, a cross-granularity evaluation protocol that disentangles assessment into three dimensions: Alignment for verifying instruction following, Fidelity for testing image quality and identity preservation, and Relative Expression Gain for quantifying the magnitude of expression changes, effectively mitigating the aforementioned evaluation biases. Third, we benchmark 18 image editing models, revealing that current approaches struggle to simultaneously achieve high fidelity and accurate expression manipulation, with fine-grained instruction following identified as the primary bottleneck. Finally, leveraging the scalable characteristic of introduced benchmark engine, we provide a 20k+ in-the-wild facial training set and demonstrate its effectiveness by fine-tuning a baseline model that achieves significant performance gains. Our benchmark and related code will be made publicly open soon.

Problem

• Facial Expression Editing
• Benchmark
• Evaluation Bias
• Disentangled Evaluation
• Fine-grained Control

Innovation

• Facial Expression Editing
• Disentangled Evaluation
• Cross-Granular Benchmark
• FED-Score
• Instruction-Following