Mutual Information Divergence: A Unified Metric for Multimodal Generative Models

25 May 2022  ·  Jin-Hwa Kim, Yunji Kim, Jiyoung Lee, Kang Min Yoo, Sang-Woo Lee ·

Text-to-image generation and image captioning are recently emerged as a new experimental paradigm to assess machine intelligence. They predict continuous quantity accompanied by their sampling techniques in the generation, making evaluation complicated and intractable to get marginal distributions. Based on a recent trend that multimodal generative evaluations exploit a vison-and-language pre-trained model, we propose the negative Gaussian cross-mutual information using the CLIP features as a unified metric, coined by Mutual Information Divergence (MID). To validate, we extensively compare it with competing metrics using carefully-generated or human-annotated judgments in text-to-image generation and image captioning tasks. The proposed MID significantly outperforms the competitive methods by having consistency across benchmarks, sample parsimony, and robustness toward the exploited CLIP model. We look forward to seeing the underrepresented implications of the Gaussian cross-mutual information in multimodal representation learning and the future works based on this novel proposition.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Human Judgment Correlation Flickr8k-CF MID Kendall's Tau-b 37.3 # 1
Human Judgment Correlation Flickr8k-Expert MID Kendall's Tau-c 54.9 # 1
Hallucination Pair-wise Detection (4-ref) FOIL MID Mean Accuracy 90.5 # 2
Hallucination Pair-wise Detection (1-ref) FOIL MID Mean Accuracy 90.5 # 2
Human Judgment Classification Pascal-50S MID Mean Accuracy 85.2 # 1