cs.CVNov 15, 2026

A Threshold-Free Visual-Semantic Matching Metric for Evaluating Consistency in Subject-Driven Image Generation

Authors: Bowei Shan, Heng Yin

Organizations: School of Information Engineering, Chang’an University, Xi’an, China

Abstract

We propose a Threshold-Free Visual-Sematic Matching (TF-VSM) metric, to quantitatively evaluate localized visual consistency across varying poses and environments in subject-driven text-to-image generation. TF-VSM is a fully continuous, threshold-free visual consistency metric paired with an extended feature-disentanglement retraining schedule for dual-branch ResNet feature-aggregation networks. Existing metrics suffer from point-alignment dropouts, threshold sensitivity, and score instability. To address them, our TF-VSM, rather than enforcing rigid binary cutoffs (TsT_s, TvT_v), weights every object point by a non-linear semantic match confidence wj=ReLU(s^j)2w_j = ReLU(\hat s_j)^2 to yield a closed-form confidence-weighted score without any hand-tuned hyperparameters. Experimental results show that, TF-VSM attains Pearson 0.638 and Spearman 0.674 oracle correlation, improving over the pretrained VSM baseline (0.454/0.589) by +0.171/ + 0.116 after retraining, and drastically outperforming CLIP (-0.052/ -0.004), DINO (0.087/0.121), and VLMs (0.096/0.003). Six diagnostic ablations confirm that smooth confidence weighting is the right mechanism, that hard semantic gating is brittle (dropping to 0.533 Pearson at TsT_s=0.8), and that the threshold-free metric is essentially invariant to the training time loss balance αα. Code is available at: https://github.com/bwshan/TF-VSM.

Figures & tables

Appendix figures & tables2 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows

    Oct 2, 2026Shuai Fu, Jing Gu, Jian Zhou +3Image Generation EvaluationT2I Generation Evaluation

  2. The Many Senses of Visual Similarity: A Text-Prompted Image Perceptual Metric

    Jul 20, 2026Sheng-Yu Wang, Yotam Nitzan, Aaron Hertzmann +4VLM EvaluationT2I Generation Evaluation

  3. Decomposing Subject-Driven Image Generation via Intermediate Structural Prediction

    May 20, 2026Hanzhong Guo, Yizhou YuGeometry-Conditioned Image GenerationSubject-Driven Image Generation