MobileCLIP2: Improving Multi-Modal Reinforced Training

• URL: http://arxiv.org/abs/2508.20691v1
• Date: Thu, 28 Aug 2025 11:50:22 GMT
• Title: MobileCLIP2: Improving Multi-Modal Reinforced Training
• Authors: Fartash Faghri, Pavan Kumar Anasosalu Vasu, Cem Koc, Vaishaal Shankar, Alexander Toshev, Oncel Tuzel, Hadi Pouransari,
• Abstract summary: We train a new family of models called MobileCLIP2 and achieve state-of-the-art ImageNet-1k zero-shot accuracies at low latencies.<n>We observe 2.2% improvement in ImageNet-1k accuracy for MobileCLIP2-B compared with MobileCLIP-B architecture.
• Score: 65.61629555586948
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Foundation image-text models such as CLIP with zero-shot capabilities enable a wide array of applications. MobileCLIP is a recent family of image-text models at 3-15ms latency and 50-150M parameters with state-of-the-art zero-shot accuracy. The main ingredients in MobileCLIP were its low-latency and light architectures and a novel multi-modal reinforced training that made knowledge distillation from multiple caption-generators and CLIP teachers efficient, scalable, and reproducible. In this paper, we improve the multi-modal reinforced training of MobileCLIP through: 1) better CLIP teacher ensembles trained on the DFN dataset, 2) improved captioner teachers trained on the DFN dataset and fine-tuned on a diverse selection of high-quality image-caption datasets. We discover new insights through ablations such as the importance of temperature tuning in contrastive knowledge distillation, the effectiveness of caption-generator fine-tuning for caption diversity, and the additive improvement from combining synthetic captions generated by multiple models. We train a new family of models called MobileCLIP2 and achieve state-of-the-art ImageNet-1k zero-shot accuracies at low latencies. In particular, we observe 2.2% improvement in ImageNet-1k accuracy for MobileCLIP2-B compared with MobileCLIP-B architecture. Notably, MobileCLIP2-S4 matches the zero-shot accuracy of SigLIP-SO400M/14 on ImageNet-1k while being 2$\times$ smaller and improves on DFN ViT-L/14 at 2.5$\times$ lower latency. We release our pretrained models (https://github.com/apple/ml-mobileclip) and the data generation code (https://github.com/apple/ml-mobileclip-dr). The data generation code makes it easy to create new reinforced datasets with arbitrary teachers using distributed scalable processing.

Related papers

• HQ-CLIP: Leveraging Large Vision-Language Models to Create High-Quality Image-Text Datasets and CLIP Models [15.877790469608662]
  We introduce an LVLM-driven data refinement pipeline to enhance the quality of image-text pair data.<n>We propose a training paradigm that extends conventional contrastive learning by incorporating negative descriptions and short tags.<n>Our approach achieves state-of-the-art performance in zero-shot classification, cross-modal retrieval, and fine-grained visual understanding tasks.
  arXiv  Detail & Related papers  (2025-07-30T07:21:36Z)
• TULIP: Towards Unified Language-Image Pretraining [60.99500935831526]
  We introduce T, an open-source, drop-in replacement for existing CLIP-like models.<n>Our method leverages generative data augmentation, enhanced image-image and text-text contrastive learning, and image/text reconstruction regularization to learn fine-grained visual features.<n>Our approach, scaling to over 1B parameters, outperforms existing state-of-the-art (SOTA) models across benchmarks.
  arXiv  Detail & Related papers  (2025-03-19T17:58:57Z)
• IDEA: Image Description Enhanced CLIP-Adapter [23.446016867479138]
  We propose an Image Description Enhanced CLIP-Adapter (IDEA) method to adapt CLIP to few-shot image classification tasks.<n>IDEA captures fine-grained features by leveraging both visual features and textual descriptions of images.<n>As one important contribution, we employ the Llama model and design a comprehensive pipeline to generate textual descriptions for images of 11 datasets.
  arXiv  Detail & Related papers  (2025-01-15T14:12:59Z)
• MobileCLIP: Fast Image-Text Models through Multi-Modal Reinforced Training [17.158498267947877]
  We introduce MobileCLIP, a new family of efficient image-text models optimized for runtime performance. MobileCLIP uses knowledge transfer from an image captioning model and an ensemble of strong CLIP encoders to improve the accuracy of efficient models. Our approach avoids train-time compute overhead by storing the additional knowledge in a reinforced dataset.
  arXiv  Detail & Related papers  (2023-11-28T18:55:42Z)
• VeCLIP: Improving CLIP Training via Visual-enriched Captions [63.547204530720705]
  This study introduces a scalable pipeline for noisy caption rewriting. We emphasize the incorporation of visual concepts into captions, termed as Visual-enriched Captions (VeCap) We showcase the adaptation of this method for training CLIP on large-scale web-crawled datasets, termed VeCLIP.
  arXiv  Detail & Related papers  (2023-10-11T17:49:13Z)
• Improving CLIP Training with Language Rewrites [57.935517901210225]
  We introduce Language augmented CLIP (LaCLIP) to enhance CLIP training through language rewrites. We show that LaCLIP significantly improves the transfer performance without computation or memory overhead during training. Specifically for ImageNet zero-shot accuracy, LaCLIP outperforms CLIP by 8.2% on CC12M and 2.4% on LAION-400M.
  arXiv  Detail & Related papers  (2023-05-31T17:59:04Z)
• Improving Zero-shot Generalization and Robustness of Multi-modal Models [70.14692320804178]
  Multi-modal image-text models such as CLIP and LiT have demonstrated impressive performance on image classification benchmarks. We investigate the reasons for this performance gap and find that many of the failure cases are caused by ambiguity in the text prompts. We propose a simple and efficient way to improve accuracy on such uncertain images by making use of the WordNet hierarchy.
  arXiv  Detail & Related papers  (2022-12-04T07:26:24Z)
• Masked Unsupervised Self-training for Zero-shot Image Classification [98.23094305347709]
  Masked Unsupervised Self-Training (MUST) is a new approach which leverages two different and complimentary sources of supervision: pseudo-labels and raw images. MUST improves upon CLIP by a large margin and narrows the performance gap between unsupervised and supervised classification.
  arXiv  Detail & Related papers  (2022-06-07T02:03:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.