VinVL: Revisiting Visual Representations in Vision-Language Models
This work provides a substantial improvement in visual feature extraction for vision-language models, benefiting researchers and practitioners working on multimodal AI tasks.
This paper tackles the problem of improving visual representations for vision-language (VL) tasks by developing a new object detection model. The new model, which is larger and pre-trained on more data, significantly improves performance across all VL tasks, achieving new state-of-the-art results on seven public benchmarks.
This paper presents a detailed study of improving visual representations for vision language (VL) tasks and develops an improved object detection model to provide object-centric representations of images. Compared to the most widely used \emph{bottom-up and top-down} model \cite{anderson2018bottom}, the new model is bigger, better-designed for VL tasks, and pre-trained on much larger training corpora that combine multiple public annotated object detection datasets. Therefore, it can generate representations of a richer collection of visual objects and concepts. While previous VL research focuses mainly on improving the vision-language fusion model and leaves the object detection model improvement untouched, we show that visual features matter significantly in VL models. In our experiments we feed the visual features generated by the new object detection model into a Transformer-based VL fusion model \oscar \cite{li2020oscar}, and utilize an improved approach \short\ to pre-train the VL model and fine-tune it on a wide range of downstream VL tasks. Our results show that the new visual features significantly improve the performance across all VL tasks, creating new state-of-the-art results on seven public benchmarks. We will release the new object detection model to public.