Advancing Image Classification with Discrete Diffusion Classification Modeling
This addresses performance issues in image classification for scenarios with data uncertainty, offering a novel method that improves over standard classifiers, though it is incremental as it builds on existing diffusion techniques.
The paper tackles image classification under high-uncertainty conditions like corrupted inputs or limited data by proposing Discrete Diffusion Classification Modeling (DiDiCM), which uses a diffusion-based procedure to model class label distributions, achieving higher accuracy on ImageNet with gains increasing with task difficulty.
Image classification is a well-studied task in computer vision, and yet it remains challenging under high-uncertainty conditions, such as when input images are corrupted or training data are limited. Conventional classification approaches typically train models to directly predict class labels from input images, but this might lead to suboptimal performance in such scenarios. To address this issue, we propose Discrete Diffusion Classification Modeling (DiDiCM), a novel framework that leverages a diffusion-based procedure to model the posterior distribution of class labels conditioned on the input image. DiDiCM supports diffusion-based predictions either on class probabilities or on discrete class labels, providing flexibility in computation and memory trade-offs. We conduct a comprehensive empirical study demonstrating the superior performance of DiDiCM over standard classifiers, showing that a few diffusion iterations achieve higher classification accuracy on the ImageNet dataset compared to baselines, with accuracy gains increasing as the task becomes more challenging. We release our code at https://github.com/omerb01/didicm .