Overcoming Language Priors for Visual Question Answering Based on Knowledge Distillation

• URL: http://arxiv.org/abs/2501.05690v1
• Date: Fri, 10 Jan 2025 03:42:37 GMT
• Title: Overcoming Language Priors for Visual Question Answering Based on Knowledge Distillation
• Authors: Daowan Peng, Wei Wei,
• Abstract summary: Previous studies have pointed out that visual question answering (VQA) models are prone to relying on language priors for answer predictions. We propose a novel method, namely, KDAR, leveraging knowledge distillation to address the prior-dependency dilemmas within the VQA task. Our method achieves state-of-the-art performance on the VQA-CPv2 out-of-distribution (OOD) benchmark.
• Score: 7.755718939237373
• License:
• Abstract: Previous studies have pointed out that visual question answering (VQA) models are prone to relying on language priors for answer predictions. In this context, predictions often depend on linguistic shortcuts rather than a comprehensive grasp of multimodal knowledge, which diminishes their generalization ability. In this paper, we propose a novel method, namely, KDAR, leveraging knowledge distillation to address the prior-dependency dilemmas within the VQA task. Specifically, the regularization effect facilitated by soft labels from a well-trained teacher is employed to penalize overfitting to the most common answers. The soft labels, which serve a regularization role, also provide semantic guidance that narrows the range of candidate answers. Additionally, we design an adaptive sample-wise reweighting learning strategy to further mitigate bias by dynamically adjusting the importance of each sample. Experimental results demonstrate that our method enhances performance in both OOD and IID settings. Our method achieves state-of-the-art performance on the VQA-CPv2 out-of-distribution (OOD) benchmark, significantly outperforming previous state-of-the-art approaches.

Related papers

• CRoFT: Robust Fine-Tuning with Concurrent Optimization for OOD Generalization and Open-Set OOD Detection [42.33618249731874]
  We show that minimizing the magnitude of energy scores on training data leads to domain-consistent Hessians of classification loss. We have developed a unified fine-tuning framework that allows for concurrent optimization of both tasks.
  arXiv  Detail & Related papers  (2024-05-26T03:28:59Z)
• Debiasing Multimodal Large Language Models [61.6896704217147]
  Large Vision-Language Models (LVLMs) have become indispensable tools in computer vision and natural language processing. Our investigation reveals a noteworthy bias in the generated content, where the output is primarily influenced by the underlying Large Language Models (LLMs) prior to the input image. To rectify these biases and redirect the model's focus toward vision information, we introduce two simple, training-free strategies.
  arXiv  Detail & Related papers  (2024-03-08T12:35:07Z)
• Robust Visual Question Answering: Datasets, Methods, and Future Challenges [23.59923999144776]
  Visual question answering requires a system to provide an accurate natural language answer given an image and a natural language question. Previous generic VQA methods often exhibit a tendency to memorize biases present in the training data rather than learning proper behaviors, such as grounding images before predicting answers. Various datasets and debiasing methods have been proposed to evaluate and enhance the VQA robustness, respectively.
  arXiv  Detail & Related papers  (2023-07-21T10:12:09Z)
• Unveiling Cross Modality Bias in Visual Question Answering: A Causal View with Possible Worlds VQA [111.41719652451701]
  We first model a confounding effect that causes language and vision bias simultaneously. We then propose a counterfactual inference to remove the influence of this effect. The proposed method outperforms the state-of-the-art methods in VQA-CP v2 datasets.
  arXiv  Detail & Related papers  (2023-05-31T09:02:58Z)
• An Empirical Study on the Language Modal in Visual Question Answering [31.692905677913068]
  Generalization beyond in-domain experience to out-of-distribution data is of paramount significance in the AI domain. This paper attempts to provide new insights into the influence of language modality on VQA performance.
  arXiv  Detail & Related papers  (2023-05-17T11:56:40Z)
• ASPEST: Bridging the Gap Between Active Learning and Selective Prediction [56.001808843574395]
  Selective prediction aims to learn a reliable model that abstains from making predictions when uncertain. Active learning aims to lower the overall labeling effort, and hence human dependence, by querying the most informative examples. In this work, we introduce a new learning paradigm, active selective prediction, which aims to query more informative samples from the shifted target domain.
  arXiv  Detail & Related papers  (2023-04-07T23:51:07Z)
• Fairness-guided Few-shot Prompting for Large Language Models [93.05624064699965]
  In-context learning can suffer from high instability due to variations in training examples, example order, and prompt formats. We introduce a metric to evaluate the predictive bias of a fixed prompt against labels or a given attributes. We propose a novel search strategy based on the greedy search to identify the near-optimal prompt for improving the performance of in-context learning.
  arXiv  Detail & Related papers  (2023-03-23T12:28:25Z)
• Introspective Distillation for Robust Question Answering [70.18644911309468]
  Question answering (QA) models are well-known to exploit data bias, e.g., the language prior in visual QA and the position bias in reading comprehension. Recent debiasing methods achieve good out-of-distribution (OOD) generalizability with a considerable sacrifice of the in-distribution (ID) performance. We present a novel debiasing method called Introspective Distillation (IntroD) to make the best of both worlds for QA.
  arXiv  Detail & Related papers  (2021-11-01T15:30:15Z)
• MUTANT: A Training Paradigm for Out-of-Distribution Generalization in Visual Question Answering [58.30291671877342]
  We present MUTANT, a training paradigm that exposes the model to perceptually similar, yet semantically distinct mutations of the input. MUTANT establishes a new state-of-the-art accuracy on VQA-CP with a $10.57%$ improvement.
  arXiv  Detail & Related papers  (2020-09-18T00:22:54Z)

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.