This paper introduces the Recognize Anything Plus Model (RAM++), an open-set image tagging model effectively leveraging multi-grained text supervision. Previous approaches (e.g., CLIP) primarily utilize global text supervision paired with images, leading to sub-optimal performance in recognizing multiple individual semantic tags. In contrast, RAM++ seamlessly integrates individual tag supervision with global text supervision, all within a unified alignment framework. This integration not only ensures efficient recognition of predefined tag categories, but also enhances generalization capabilities for diverse open-set categories. Furthermore, RAM++ employs large language models (LLMs) to convert semantically constrained tag supervision into more expansive tag description supervision, thereby enriching the scope of open-set visual description concepts. Comprehensive evaluations on various image recognition benchmarks demonstrate RAM++ exceeds existing state-of-the-art (SOTA) open-set image tagging models on most aspects. Specifically, for predefined commonly used tag categories, RAM++ showcases 10.2 mAP and 15.4 mAP enhancements over CLIP on OpenImages and ImageNet. For open-set categories beyond predefined, RAM++ records improvements of 5.0 mAP and 6.4 mAP over CLIP and RAM respectively on OpenImages. For diverse human-object interaction phrases, RAM++ achieves 7.8 mAP and 4.7 mAP improvements on the HICO benchmark.

This paper presents Tag2Text, a vision language pre-training (VLP) framework, which introduces image tagging into vision-language models to guide the learning of visual-linguistic features. In contrast to prior works which utilize object tags either manually labeled or automatically detected with a limited detector, our approach utilizes tags parsed from its paired text to learn an image tagger and meanwhile provides guidance to vision-language models. Given that, Tag2Text can utilize large-scale annotation-free image tags in accordance with image-text pairs, and provides more diverse tag categories beyond objects. Strikingly, Tag2Text showcases the ability of a foundational image tagging model, with superior zero-shot performance even comparable to full supervision manner. Moreover, by leveraging tagging guidance, Tag2Text effectively enhances the performance of vision-language models on both generation-based and alignment-based tasks. Across a wide range of downstream benchmarks, Tag2Text achieves state-of-the-art results with similar model sizes and data scales, demonstrating the efficacy of the proposed tagging guidance.

It is a consensus that small models perform quite poorly under the paradigm of self-supervised contrastive learning. Existing methods usually adopt a large off-the-shelf model to transfer knowledge to the small one via distillation. Despite their effectiveness, distillation-based methods may not be suitable for some resource-restricted scenarios due to the huge computational expenses of deploying a large model. In this paper, we study the issue of training self-supervised small models without distillation signals. We first evaluate the representation spaces of the small models and make two non-negligible observations: (i) the small models can complete the pretext task without overfitting despite their limited capacity and (ii) they universally suffer the problem of over clustering. Then we verify multiple assumptions that are considered to alleviate the over-clustering phenomenon. Finally, we combine the validated techniques and improve the baseline performances of five small architectures with considerable margins, which indicates that training small self-supervised contrastive models is feasible even without distillation signals. The code is available at https://github.com/WOWNICE/ssl-small.

Vision-Language Pre-training (VLP) with large-scale image-text pairs has demonstrated superior performance in various fields. However, the image-text pairs co-occurrent on the Internet typically lack explicit alignment information, which is suboptimal for VLP. Existing methods proposed to adopt an off-the-shelf object detector to utilize additional image tag information. However, the object detector is time-consuming and can only identify the pre-defined object categories, limiting the model capacity. Inspired by the observation that the texts incorporate incomplete fine-grained image information, we introduce IDEA, which stands for increasing text diversity via online multi-label recognition for VLP. IDEA shows that multi-label learning with image tags extracted from the texts can be jointly optimized during VLP. Moreover, IDEA can identify valuable image tags online to provide more explicit textual supervision. Comprehensive experiments demonstrate that IDEA can significantly boost the performance on multiple downstream datasets with a small extra computational cost.

Metric learning with neural networks has exhibited promising improvements in representation learning. Yet cross-modal retrieval poses a unique challenge to metric learning: how to compute the distance across different modalities such as image and text. Existing neural network based methods tend to establish two branches for images and texts respectively to bridge the modal gap. Also, most of them cannot fully exploit the structure embedded in the multimodal data. This paper introduces embedding layer to provide cross-modal shared representation with non-linearity and reformulates the cross-modal retrieval problem as a graph embedding problem by constructing a multimodal graph. To learn the graph embedding, training pairs and triplets are uniformly generated from random walk sequences on the graph. Then graph pair and triplet constraints are imposed on the embedding layer for structure preservation. Meanwhile, a classifier is trained with labeled data to ensure the learned embedding is coupled with semantic information. For optimization, graph pair and triplet constraints are integrated into a unified multi-task learning with the supervised classifier. Experimental results on the Wiki and NUS-WIDE datasets demonstrate the effectiveness and superiority of the learned embedding for cross-modal retrieval.