Pretrained Reversible Generation as Unsupervised Visual Representation Learning
This addresses the challenge of reusing high-capacity generative models for robust feature extraction in computer vision, representing an incremental improvement over prior generative-based methods.
The paper tackles the problem of leveraging generative models for discriminative tasks by proposing Pretrained Reversible Generation (PRG), which extracts unsupervised representations by reversing a pretrained generative process, achieving state-of-the-art performance with 78% top-1 accuracy on ImageNet at 64*64 resolution.
Recent generative models based on score matching and flow matching have significantly advanced generation tasks, but their potential in discriminative tasks remains underexplored. Previous approaches, such as generative classifiers, have not fully leveraged the capabilities of these models for discriminative tasks due to their intricate designs. We propose Pretrained Reversible Generation (PRG), which extracts unsupervised representations by reversing the generative process of a pretrained continuous generation model. PRG effectively reuses unsupervised generative models, leveraging their high capacity to serve as robust and generalizable feature extractors for downstream tasks. This framework enables the flexible selection of feature hierarchies tailored to specific downstream tasks. Our method consistently outperforms prior approaches across multiple benchmarks, achieving state-of-the-art performance among generative model based methods, including 78% top-1 accuracy on ImageNet at a resolution of 64*64. Extensive ablation studies, including out-of-distribution evaluations, further validate the effectiveness of our approach.PRG is available at https://github.com/opendilab/PRG.