RA-ClipScore: Making Generative Model Evaluation More Interpretable

• URL: http://arxiv.org/abs/2608.12088v1
• Date: Wed, 12 Aug 2026 14:11:27 GMT
• Title: RA-ClipScore: Making Generative Model Evaluation More Interpretable
• Abstract summary: We propose RA-CLIPScore, a novel metric for semantic evaluation of generative models.<n> RA-CLIPScore measures whether generated objects adhere to the positional priors found in the training data.<n>We show that RA-CLIPScore provides more robust and interpretable evaluations than prior methods.
• Score: 10.211263272517911
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Generative models can produce images nearly indistinguishable from real data, yet rigorous and interpretable evaluation remains challenging. Conventional metrics such as FID provide only scalar scores with limited diagnostic insight. Widely adopted CLIP-based metrics enable semantic evaluation beyond simple training class labels, but inherit limitations from CLIP's training paradigm that restrict attribute-wise analysis. We propose RA-CLIPScore, a novel metric that mitigates these issues and extends CLIP-based evaluation to spatial distribution alignment, measuring whether generated objects adhere to the positional priors found in the training data. RA-CLIPScore introduces dual prompts to decouple competing attributes and leverages local patch tokens to capture fine-grained regional semantics. We evaluate image generative models on their ability to match both attribute and spatial distributions of the training data. Extensive experiments show that RA-CLIPScore provides more robust and interpretable evaluations than prior methods, particularly under distribution misalignment or partially irrelevant textual attributes. We further demonstrate how it reveals spatial biases in generative models. User evaluations confirm that Regional Single Attribute Divergence based on our RA-CLIPScore aligns more closely with human perception of visual diversity than existing semantic metrics.

Related papers

• Structuring Semantic Embeddings for Principle Evaluation: A Prototype-Guided Contrastive Learning Approach [58.15427952222424]
  General-purpose text embeddings are widely deployed for such tasks, but broad semantic similarity can place semantically similar yet task-distinct examples.<n>We introduce Prototype-Guided Contrastive Learning (PGCL), a prototype-guided geometric regularization module built on top of frozen text embeddings.
  arXiv  Detail & Related papers  (2026-08-15T13:19:25Z)
• Bridging the Semantic Gap for Categorical Data Clustering via Large Language Models [64.58262227709842]
  ARISE (Attention-weighted Representation with Integrated Semantic Embeddings) is presented.<n>It builds semantic-aware representations that complement the metric space of categorical data for accurate clustering.<n>Experiments on eight benchmark datasets demonstrate consistent improvements over seven representative counterparts.
  arXiv  Detail & Related papers  (2026-01-03T11:37:46Z)
• Structure-Aware Feature Rectification with Region Adjacency Graphs for Training-Free Open-Vocabulary Semantic Segmentation [22.409969687852506]
  We propose a structure-aware feature rectification approach that incorporates instance-specific priors derived directly from the image.<n>Our method effectively suppresses segmentation noise, improves region-level consistency, and achieves strong performance on multiple open-vocabulary segmentation benchmarks.
  arXiv  Detail & Related papers  (2025-12-08T10:00:36Z)
• SegDebias: Test-Time Bias Mitigation for ViT-Based CLIP via Segmentation [23.565073829744676]
  We propose a test-time debiasing method for ViT based CLIP models that requires no additional training or assumptions of bias annotations.<n>Our approach uses a pretrained segmentation model to isolate the target visual attribute, then adjusts the non target regions so that their embeddings are uniformly similar to all class specific text prompts.<n>Experiments on Waterbirds and CelebA show that our method outperforms existing test-time debiasing approaches in both group robustness metrics and Attention IoU.
  arXiv  Detail & Related papers  (2025-11-01T11:57:57Z)
• Latent Feature Alignment: Discovering Biased and Interpretable Subpopulations in Face Recognition Models [2.098253213160396]
  Latent Feature Alignment (LFA) is an attribute-label-free algorithm that uses latent directions to identify subpopulations.<n>LFA consistently outperforms k-means and nearest-neighbor search in intra-group semantic coherence.<n>These results position LFA as a practical method for representation auditing of face recognition models.
  arXiv  Detail & Related papers  (2025-10-17T10:49:50Z)
• Helping CLIP See Both the Forest and the Trees: A Decomposition and Description Approach [43.419607730361996]
  Vision-Language Models (VLMs) like CLIP achieve cross-modal alignment through contrastive learning.<n>Traditional prompt engineering relies on coarse-grained category labels, neglecting fine-grained local semantics.<n>We propose a plug-and-play solution that enables CLIP to process localized visual descriptors.
  arXiv  Detail & Related papers  (2025-07-04T10:24:26Z)
• Downstream-Pretext Domain Knowledge Traceback for Active Learning [138.02530777915362]
  We propose a downstream-pretext domain knowledge traceback (DOKT) method that traces the data interactions of downstream knowledge and pre-training guidance. DOKT consists of a traceback diversity indicator and a domain-based uncertainty estimator. Experiments conducted on ten datasets show that our model outperforms other state-of-the-art methods.
  arXiv  Detail & Related papers  (2024-07-20T01:34:13Z)
• A Closer Look at Benchmarking Self-Supervised Pre-training with Image Classification [51.35500308126506]
  Self-supervised learning (SSL) is a machine learning approach where the data itself provides supervision, eliminating the need for external labels. We study how classification-based evaluation protocols for SSL correlate and how well they predict downstream performance on different dataset types.
  arXiv  Detail & Related papers  (2024-07-16T23:17:36Z)
• Leveraging vision-language models for fair facial attribute classification [19.93324644519412]
  General-purpose vision-language model (VLM) is a rich knowledge source for common sensitive attributes. We analyze the correspondence between VLM predicted and human defined sensitive attribute distribution. Experiments on multiple benchmark facial attribute classification datasets show fairness gains of the model over existing unsupervised baselines.
  arXiv  Detail & Related papers  (2024-03-15T18:37:15Z)
• Learning disentangled representations for explainable chest X-ray classification using Dirichlet VAEs [68.73427163074015]
  This study explores the use of the Dirichlet Variational Autoencoder (DirVAE) for learning disentangled latent representations of chest X-ray (CXR) images. The predictive capacity of multi-modal latent representations learned by DirVAE models is investigated through implementation of an auxiliary multi-label classification task.
  arXiv  Detail & Related papers  (2023-02-06T18:10:08Z)
• Using Representation Expressiveness and Learnability to Evaluate Self-Supervised Learning Methods [61.49061000562676]
  We introduce Cluster Learnability (CL) to assess learnability. CL is measured in terms of the performance of a KNN trained to predict labels obtained by clustering the representations with K-means. We find that CL better correlates with in-distribution model performance than other competing recent evaluation schemes.
  arXiv  Detail & Related papers  (2022-06-02T19:05:13Z)
• Explaining Predictions by Approximating the Local Decision Boundary [3.60160227126201]
  We present a new procedure for local decision boundary approximation (DBA) We train a variational autoencoder to learn a Euclidean latent space of encoded data representations. We exploit attribute annotations to map the latent space to attributes that are meaningful to the user.
  arXiv  Detail & Related papers  (2020-06-14T19:12:42Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.