2000 character limit reached
Faithful and Plausible Natural Language Explanations for Image Classification: A Pipeline Approach (2407.20899v3)
Published 30 Jul 2024 in cs.AI and cs.CL
Abstract: Existing explanation methods for image classification struggle to provide faithful and plausible explanations. This paper addresses this issue by proposing a post-hoc natural language explanation method that can be applied to any CNN-based classifier without altering its training process or affecting predictive performance. By analysing influential neurons and the corresponding activation maps, the method generates a faithful description of the classifier's decision process in the form of a structured meaning representation, which is then converted into text by a LLM. Through this pipeline approach, the generated explanations are grounded in the neural network architecture, providing accurate insight into the classification process while remaining accessible to non-experts. Experimental results show that the NLEs constructed by our method are significantly more plausible and faithful. In particular, user interventions in the neural network structure (masking of neurons) are three times more effective than the baselines.
- Gpt-4 technical report.
- Faithfulness tests for natural language explanations. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 283–294, Toronto, Canada. Association for Computational Linguistics.
- On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation. PLoS ONE, 10.
- Network dissection: Quantifying interpretability of deep visual representations.
- e-snli: Natural language inference with natural language explanations. In S. Bengio, H. Wallach, H. Larochelle, K. Grauman, N. Cesa-Bianchi, and R. Garnett, editors, Advances in Neural Information Processing Systems 31, pages 9539–9549. Curran Associates, Inc.
- This looks like that: deep learning for interpretable image recognition. Advances in neural information processing systems, 32.
- Janez Demšar. 2006. Statistical comparisons of classifiers over multiple data sets. Journal of Machine Learning Research, 7(1):1–30.
- Accountability of ai under the law: The role of explanation. arXiv preprint arXiv:1711.01134.
- Rationalization for explainable nlp: A survey. Frontiers in Artificial Intelligence, 6.
- Deep residual learning for image recognition. CoRR, abs/1512.03385.
- Generating visual explanations. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part IV 14, pages 3–19. Springer.
- Natural language descriptions of deep visual features. In International Conference on Learning Representations.
- Alon Jacovi and Yoav Goldberg. 2020. Towards faithfully interpretable NLP systems: How should we define and evaluate faithfulness? In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 4198–4205, Online. Association for Computational Linguistics.
- Vidhya Kamakshi and Narayanan C. Krishnan. 2023. Explainable image classification: The journey so far and the road ahead. AI, 4(3):620–651.
- Explaining chest x-ray pathologies in natural language. In Medical Image Computing and Computer Assisted Intervention – MICCAI 2022, pages 701–713, Cham. Springer Nature Switzerland.
- Textual explanations for self-driving vehicles. In Computer Vision – ECCV 2018, pages 577–593, Cham. Springer International Publishing.
- Captum: A unified and generic model interpretability library for pytorch.
- Alon Lavie and Abhaya Agarwal. 2007. METEOR: An automatic metric for MT evaluation with high levels of correlation with human judgments. In Proceedings of the Second Workshop on Statistical Machine Translation, pages 228–231, Prague, Czech Republic. Association for Computational Linguistics.
- Scott M Lundberg and Su-In Lee. 2017. A unified approach to interpreting model predictions. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems 30, pages 4765–4774. Curran Associates, Inc.
- Natural language rationales with full-stack visual reasoning: From pixels to semantic frames to commonsense graphs. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 2810–2829, Online. Association for Computational Linguistics.
- Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311–318, Philadelphia, Pennsylvania, USA. Association for Computational Linguistics.
- Rise: Randomized input sampling for explanation of black-box models.
- The problem of coherence in natural language explanations of recommendations. arXiv preprint arXiv:2312.11356.
- " why should i trust you?" explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, pages 1135–1144.
- "why should I trust you?": Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, August 13-17, 2016, pages 1135–1144.
- Nlx-gpt: A model for natural language explanations in vision and vision-language tasks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8322–8332.
- Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE international conference on computer vision, pages 618–626.
- Grad-cam: Visual explanations from deep networks via gradient-based localization. In 2017 IEEE International Conference on Computer Vision (ICCV), pages 618–626.
- Explainable image classification with evidence counterfactual. Pattern Analysis and Applications, 25(2):315–335.
- Comprehensible convolutional neural networks via guided concept learning. In 2021 International Joint Conference on Neural Networks (IJCNN), pages 1–8.
- Measuring association between labels and free-text rationales. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 10266–10284, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics.
- Jialin Wu and Raymond Mooney. 2019. Faithful multimodal explanation for visual question answering. In Proceedings of the 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP, pages 103–112, Florence, Italy. Association for Computational Linguistics.
- Show, attend and tell: Neural image caption generation with visual attention. In International conference on machine learning, pages 2048–2057. PMLR.