A Fistful of Words: Learning Transferable Visual Models from Bag-of-Words Supervision

By Ajinkya Tejankar, Bichen Wu, Saining Xie, Madian Khabsa, Hamed Pirsiavash, Hamed Firooz. 2021

23 citations recorded by the source corpus.

Using natural language as a supervision for training visual recognition models holds great promise. Recent works have shown that if such supervision is used in the form of alignment between images and captions in large training datasets, then the resulting aligned models perform well on zero-shot classification as downstream tasks2. In this paper, we focus on teasing out what parts of the language supervision are essential for training zero-shot image classification models. Through extensive and careful experiments, we show that: 1) A simple Bag-of-Words (BoW) caption could be used as a replacement for most of the image captions in the dataset. Surprisingly, we observe that this approach improves the zero-shot classification performance when combined with word balancing. 2) Using a BoW pretrained model, we can obtain more training data by generating pseudo-BoW captions on images that do not have a caption. Models trained on images with real and pseudo-BoW captions achieve stronger zero-shot performance. On ImageNet-1k zero-shot evaluation, our best model, that uses only 3M image-caption pairs, performs on-par with a CLIP model trained on 15M image-caption pairs (31.5% vs 31.3%).

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

Citations in the corpus

When and why vision-language models behave like bags-of-words, and what to do about it? (2022, unknown): “Closer to our experiments, the work of Tejankar et al. (2021) shows that training contrastive vision language models using only bag-of-words in place of the caption does not signiﬁcantly hurt performance on zero-shot classiﬁcation.”