Revisiting Weakly Supervised Pre-Training of Visual Perception Models
This work addresses the development of visual recognition systems by demonstrating the effectiveness of weakly supervised pre-training, which is incremental as it builds on existing weakly supervised approaches with updated methods and data.
The paper tackled the problem of model pre-training for visual recognition by revisiting weakly supervised pre-training using hashtag supervision with modern residual networks and a large dataset, finding that these models are very competitive across transfer-learning settings and substantially outperform self-supervised counterparts.
Model pre-training is a cornerstone of modern visual recognition systems. Although fully supervised pre-training on datasets like ImageNet is still the de-facto standard, recent studies suggest that large-scale weakly supervised pre-training can outperform fully supervised approaches. This paper revisits weakly-supervised pre-training of models using hashtag supervision with modern versions of residual networks and the largest-ever dataset of images and corresponding hashtags. We study the performance of the resulting models in various transfer-learning settings including zero-shot transfer. We also compare our models with those obtained via large-scale self-supervised learning. We find our weakly-supervised models to be very competitive across all settings, and find they substantially outperform their self-supervised counterparts. We also include an investigation into whether our models learned potentially troubling associations or stereotypes. Overall, our results provide a compelling argument for the use of weakly supervised learning in the development of visual recognition systems. Our models, Supervised Weakly through hashtAGs (SWAG), are available publicly.