CatLIP: CLIP-level Visual Recognition Accuracy with 2.7x Faster Pre-training on Web-scale Image-Text Data
This addresses the computational cost problem for researchers and practitioners training large vision models on web-scale data, though it is an incremental improvement over existing contrastive methods.
The paper tackles the computational bottleneck of pairwise similarity in contrastive learning for vision models by reframing pre-training as a classification task, achieving 2.7x faster training on web-scale image-text data while maintaining high accuracy comparable to CLIP.
Contrastive learning has emerged as a transformative method for learning effective visual representations through the alignment of image and text embeddings. However, pairwise similarity computation in contrastive loss between image and text pairs poses computational challenges. This paper presents a novel weakly supervised pre-training of vision models on web-scale image-text data. The proposed method reframes pre-training on image-text data as a classification task. Consequently, it eliminates the need for pairwise similarity computations in contrastive loss, achieving a remarkable $2.7\times$ acceleration in training speed compared to contrastive learning on web-scale data. Through extensive experiments spanning diverse vision tasks, including detection and segmentation, we demonstrate that the proposed method maintains high representation quality. Our source code along with pre-trained model weights and training recipes is available at \url{https://github.com/apple/corenet}.