mlsanityData to 14 Oct 2026

CVE-SAI: Counterfactual Visual Evidence-Guided Selective Attribute Indexing for Risk-Controlled E-commerce Search

arXiv 2608.25023 ↗First posted 25 Aug 2026cs.AI cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Multimodal product models can complete missing e-commerce attributes, yet current methods still optimize attribute-answer accuracy without verifying visual support, conflate transient prediction with persistent index admission, and lack explicit risk control over factually incorrect or visually unsupported values. We address these gaps with Counterfactual Visual Evidence-Guided Selective Attribute Indexing (CVE-SAI), which first infers and freezes an ontology-constrained candidate from the primary image and attribute question without catalog text, and then decides whether that candidate should enter the index. Focus-Zone Distortion (FZD) constructs an attribute-specific visual-dependence proxy through a controlled counterfactual intervention, and Evidence-Guided Attention Redistribution (EGAR) uses the proxy to refine ontology-constrained scoring. The canonical candidate is frozen before evidence necessity, evidence retention, nuisance-transformation stability, and candidate-specific catalog-text conflict audits; catalog text can only tighten admission and cannot revise the candidate. Independent family-level calibration selects one policy with a simultaneous one-sided finite-sample bound under a 5% unsafe-admission budget. Experiments on five visual attributes derived from Amazon Berkeley Objects show that CVE-SAI improves attribute inference and evidence localization, achieves the highest certified admission coverage under the shared risk protocol, and yields the strongest controlled retrieval performance with the lowest unsafe auto-induced exposure among automatic-admission systems. Separating inference from admission therefore enables visually supported attribute completion to improve retrieval while limiting persistent index contamination.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 32 of its 48 references that are in the map. First posted on arXiv 25 Aug 2026.

Builds on

32 of 48 references.

References · 32

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  3. 2025Qwen2.5-VL Technical Reportcited by 5,587
  4. 2017On Calibration of Modern Neural Networkscited by 4,960
  5. 2023Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detectioncited by 3,332
  6. 2025InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Modelscited by 1,834
  7. 2023Evaluating Object Hallucination in Large Vision-Language Modelscited by 1,729
  8. 2025SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Featurescited by 1,188
  9. 2020Quantifying Attention Flow in Transformerscited by 775
  10. 2023Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decodingcited by 688
  11. 2021ABO: Dataset and Benchmarks for Real-World 3D Object Understandingcited by 366
  12. 2019SelectiveNet: A Deep Neural Network with an Integrated Reject Optioncited by 350
20 more
  1. 2022Conformal Risk Controlcited by 299
  2. 2021Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Controlcited by 271
  3. 2024Multi-Modal Hallucination Control by Visual Information Groundingcited by 208
  4. 2024MM-Embed: Universal Multimodal Retrieval with Multimodal LLMscited by 161
  5. 2024Mitigating Object Hallucinations in Large Vision-Language Models with Assembly of Global and Local Attentioncited by 74
  6. 2021MAVE: A Product Dataset for Multi-source Attribute Value Extractioncited by 53
  7. 2022Contrastive language and vision learning of general fashion conceptscited by 42
  8. 2025Cross-Modal Attention Calibration for LVLM Hallucination Mitigationcited by 23
  9. 2024ImplicitAVE: An Open-Source Dataset and Multimodal LLMs Benchmark for Implicit Attribute Value Extractioncited by 17
  10. 2024EIVEN: Efficient Implicit Attribute Value Extraction using Multimodal LLMcited by 12
  11. 2023Large Scale Generative Multimodal Attribute Extraction for E-commerce Attributescited by 11
  12. 2024Multi-Label Zero-Shot Product Attribute-Value Extractioncited by 7
  13. 2025MOON: Generative MLLM-based Multimodal Representation Learning for E-commerce Product Understandingcited by 7
  14. 2024Two-stage Risk Control with Application to Ranked Retrievalcited by 6
  15. 2025TACLR: A Scalable and Efficient Retrieval-based Method for Industrial Product Attribute Value Identificationcited by 6
  16. 2026Same Attention, Different Truths: Put Logit-Lens over Visual Attention to Detect and Mitigate LVLM Object Hallucinationcited by 4
  17. 2025Visual Zero-Shot E-Commerce Product Attribute Value Extractioncited by 3
  18. 2024Unsupervised dense retrieval with conterfactual contrastive learningcited by 2
  19. 2025Aligning Extraction and Generation for Robust Retrieval-Augmented Generationcited by 2
  20. 2025HADSF: Aspect Aware Semantic Control for Explainable Recommendationcited by 2

Built on by

Not cited yet.