Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks
This addresses the challenge of aligning image and text semantics for vision-language tasks, offering a novel approach that improves performance across multiple benchmarks.
The paper tackles the problem of learning cross-modal representations for vision-language tasks by proposing Oscar, which uses object tags as anchor points to ease alignment learning, achieving new state-of-the-art results on six tasks.
Large-scale pre-training methods of learning cross-modal representations on image-text pairs are becoming popular for vision-language tasks. While existing methods simply concatenate image region features and text features as input to the model to be pre-trained and use self-attention to learn image-text semantic alignments in a brute force manner, in this paper, we propose a new learning method Oscar (Object-Semantics Aligned Pre-training), which uses object tags detected in images as anchor points to significantly ease the learning of alignments. Our method is motivated by the observation that the salient objects in an image can be accurately detected, and are often mentioned in the paired text. We pre-train an Oscar model on the public corpus of 6.5 million text-image pairs, and fine-tune it on downstream tasks, creating new state-of-the-arts on six well-established vision-language understanding and generation tasks.