Meta tags:
description=TULIP: Towards Unified Language-Image Pretraining;
keywords=TULIP, Vision-Language, Image-Text Learning, AI, Machine Learning;

Headings (most frequently used words):

tulip, image, towards, unified, language, pretraining, abstract, enhancing, text, contrastive, learning, experimental, results, bibtex,

Text of the page (most frequently used words):
and (15), tulip (10), image (9), language (8), the (7), vision (7), contrastive (5), text (5), this (4), page (4), models (4), #towards (3), unified (3), pretraining (3), fine (3), learning (3), was (2), project (2), you (2), website (2), zineng (2), tang (2), long (2), lian (2), seun (2), eisape (2), xudong (2), wang (2), roei (2), herzig (2), adam (2), yala (2), alane (2), suhr (2), trevor (2), darrell (2), david (2), chan (2), berkeley (2), state (2), art (2), across (2), multiple (2), benchmarks (2), significantly (2), zero (2), shot (2), classification (2), grained (2), reasoning (2), existing (2), tasks (2), generative (2), data (2), augmentation (2), enhanced (2), improve (2), understanding (2), integrating (2), reconstruction (2), regularization (2), alignment (2), built, using, which, adopted, from, are, free, borrow, just, ask, that, link, back, footer, licensed, under, icons, created, mihimihi, flaticon, creative, commons, attribution, sharealike, international, license, nerfies, academic, template, misc, tang2025tulip, title, author, institution, university, california, year, 2025, note, preprint, bibtex, achieves, performance, improves, imagenet, enhances, object, recognition, boosts, multimodal, scores, compared, methods, shows, improvement, mmvp, boost, tuned, experimental, results, leverages, techniques, while, maintaining, strong, grounding, objectives, alongside, ensures, robust, enhancing, despite, success, like, clip, they, struggle, with, centric, requiring, high, fidelity, introduce, novel, model, our, approach, outperforms, setting, new, abstract, coming, soon, code, arxiv, paper,

Text of the page (random words):
tulip towards unified language image pretraining tulip towards unified language image pretraining zineng tang long lian seun eisape xudong wang roei herzig adam yala alane suhr trevor darrell david chan uc berkeley paper arxiv code models hf models coming soon abstract despite the success of contrastive image text models like clip they struggle with vision centric tasks requiring high fidelity understanding we introduce tulip a novel model integrating generative data augmentation enhanced contrastive learning and reconstruction regularization to improve vision language alignment our approach significantly outperforms existing models across multiple benchmarks setting a new state of the art in zero shot classification and vision language reasoning tulip enhancing image text contrastive learning tulip leverages generative data augmentation and enhanced contrastive learning techniques to improve fine grained image understanding while maintaining strong language grounding by integrating image image and text text contrastive objectives alongside image text reconstruction regularization tulip ensures robust vision language alignment experimental results tulip achieves state of the art performance across multiple vision and vision language benchmarks it significantly improves zero shot classification on imagenet 1k enhances fine grained object recognition and boosts multimodal reasoning scores compared to existing methods tulip shows up to a 3 improvement on mmvp and a 2 boost in fine tuned vision tasks bibtex misc tang2025tulip title tulip towards unified language image pretraining author zineng tang and long lian and seun eisape and xudong wang and roei herzig and adam yala and alane suhr and trevor darrell and david m chan institution university of california berkeley year 2025 note preprint this page was built using the academic project page template which was adopted from the nerfies project page you are free to borrow the of this website we just ask that you link back to this page in the footer this website is licensed under a creative commons attribution sharealike 4 0 international license tulip icons created by mihimihi flaticon