Zero-Shot Product Attribute Labeling with Vision-Language Models: A Three-Tier Evaluation Framework

TL;DR

This paper introduces a three-tier evaluation framework for zero-shot fashion attribute prediction using vision-language models, highlighting their strengths and weaknesses in applicability detection and fine-grained classification.

Contribution

It proposes a systematic three-tier evaluation method and benchmarks various VLMs on fashion attribute prediction, revealing key performance insights and practical deployment considerations.

Findings

Zero-shot VLMs achieve 64.0% macro-F1, outperforming logistic regression.

VLMs excel at fine-grained classification with 70.8% F1.

Efficient models reach over 90% of flagship performance at lower costs.

Abstract

Fine-grained attribute prediction is essential for fashion retail applications including catalog enrichment, visual search, and recommendation systems. Vision-Language Models (VLMs) offer zero-shot prediction without task-specific training, yet their systematic evaluation on multi-attribute fashion tasks remains underexplored. A key challenge is that fashion attributes are often conditional. For example, "outer fabric" is undefined when no outer garment is visible. This requires models to detect attribute applicability before attempting classification. We introduce a three-tier evaluation framework that decomposes this challenge: (1) overall task performance across all classes (including NA class: suggesting attribute is not applicable) for all attributes, (2) attribute applicability detection, and (3) fine-grained classification when attributes are determinable. Using DeepFashion-MultiModal, which explicitly defines NA (meaning attribute doesn't exist or is not visible) within attribute label spaces, we benchmark nine VLMs spanning flagship (GPT-5, Gemini 2.5 Pro), efficient (GPT-5 Mini, Gemini 2.5 Flash), and ultra-efficient tiers (GPT-5 Nano, Gemini 2.5 Flash-Lite) against classifiers trained on pretrained Fashion-CLIP embeddings on 5,000 images across 18 attributes. Our findings reveal that: (1) zero-shot VLMs achieve 64.0% macro-F1, a threefold improvement over logistic regression on pretrained Fashion-CLIP embeddings; (2) VLMs excel at fine-grained classification (Tier 3: 70.8% F1) but struggle with applicability detection (Tier 2: 34.1% NA-F1), identifying a key bottleneck; (3) efficient models achieve over 90% of flagship performance at lower cost, offering practical deployment paths. This diagnostic framework enables practitioners to pinpoint whether errors stem from visibility detection or classification, guiding targeted improvements for production systems.