TIPSv2: Advancing Vision-Language Pretraining with Enhanced Patch-Text Alignment
Recent progress in vision-language pretraining has enabled significant improvements to many downstream computer vision applications, such as classification, retrieval, segmentation and depth prediction. However, a fundamental capability that these models still struggle with is aligning dense patch representations with text embeddings of corresponding concepts. In this work, we investigate this critical issue and propose novel techniques to enhance this capability in foundational vision-language models. First, we reveal that a patch-level distillation procedure significantly boosts dense patch-text alignment -- surprisingly, the patch-text alignment of the distilled student model strongly surpasses that of the teacher model. This observation inspires us to consider modifications to pretraining recipes, leading us to propose iBOT++, an upgrade to the commonly-used iBOT masked image objective, where unmasked tokens also contribute directly to the loss. This dramatically enhances patch-text alignment of pretrained models. Additionally, to improve vision-language pretraining efficiency and effectiveness, we modify the exponential moving average setup in the learning recipe, and introduce a caption sampling strategy to benefit from synthetic captions at different granularities. Combining these components, we develop TIPSv2, a new family of image-text encoder models suitable for a wide range of downstream applications. Through comprehensive experiments on 9 tasks and 20 datasets, we demonstrate strong performance, generally on par with or better than recent vision encoder models. Code and models are released via our project page at https://gdm-tipsv2.github.io/ .
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
MAKE: Multi-Aspect Knowledge-Enhanced Vision-Language Pretraining for Zero-shot Dermatological Assessment
Dermatological diagnosis represents a complex multimodal challenge that requires integrating visual features with specialized clinical knowledge. While vision-language pretraining (VLP) has advanced medical AI, its effec…
Clinical KnowledgeContrastive LearningCross-Modal RetrievalImproving Medical Visual Representations via Radiology Report Generation
Vision-language pretraining has been shown to produce high-quality visual encoders which transfer efficiently to downstream computer vision tasks. Contrastive learning approaches have increasingly been adopted for medica…
Contrastive LearningDecoderImage CaptioningMedical Image AnalysisBeyond Language Modeling: An Exploration of Multimodal Pretraining
The visual world offers a critical axis for advancing foundation models beyond language. Despite growing interest in this direction, the design space for native multimodal models remains opaque. We provide empirical clar…
M2-Encoder: Advancing Bilingual Image-Text Understanding by Large-scale Efficient Pretraining
Vision-language foundation models like CLIP have revolutionized the field of artificial intelligence. Nevertheless, VLM models supporting multi-language, e.g., in both Chinese and English, have lagged due to the relative…
GPUzero-shot-classificationZero-Shot Cross-Modal RetrievalZero-shot Image Retrieval+3Supervised Fine-tuning in turn Improves Visual Foundation Models
Image-text training like CLIP has dominated the pretraining of vision foundation models in recent years. Subsequent efforts have been made to introduce region-level visual learning into CLIP's pretraining but face scalab…