Representation alignment with pretrained vision models has recently shown strong potential for accelerating diffusion transformer training. By aligning intermediate diffusion features with clean-image representations from self-supervised vision encoders, existing methods improve convergence and generation quality. However, such alignment also introduces a non-trivial constraint: diffusion models operate on noisy inputs whose usable information varies across timesteps, while the reference features are extracted from clean images. In this paper, we revisit this mismatch from a token-level perspective. We find that, under full-token representation alignment, tokens with large alignment-gradient norms exhibit a stable spatial preference, suggesting that the alignment objective does not affect all tokens uniformly and may encourage the model to rely on the complete set of clean-image tokens. To address this issue, we propose MaskAlign, a token-subset representation alignment method that applies alignment to randomly sampled token subsets during training. By exposing the model to different token subsets across iterations, MaskAlign reduces the dependence of representation alignment on the complete token set and encourages alignment behavior that is more stable under token-subset perturbations. To mitigate the information loss caused by directly dropping tokens, we further introduce a lightweight pre-mask token mixing block that shares information across tokens before masking.
Shenzhen Institutes of Advanced Technology, Chinese Academy of Science（中国科学院深圳先进技术研究院）；China Telecom Corporation Ltd. Data&AI Technology Company（中国电信股份有限公司数据与AI科技公司）；University of Rochester（罗切斯特大学）；University of Science Technology of China（中国科学技术大学）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）
Masked Autoencoders (MAE) have been prevailing paradigms for large-scale vision representation pre-training. By reconstructing masked image patches from a small portion of visible image regions, MAE forces the model to infer semantic correlation within an image. Recently, some approaches apply semantic-rich teacher models to extract image features as the reconstruction target, leading to better performance. However, unlike the low-level features such as pixel values, we argue the features extracted by powerful teacher models already encode rich semantic correlation across regions in an intact …
Existing text-guided image manipulation methods aim to modify the appearance of the image or to edit a few objects in a virtual or simple scenario, which is far from practical application. In this work, we study a novel task on text-guided image manipulation on the entity level in the real world. The task imposes three basic requirements, (1) to edit the entity consistent with the text descriptions, (2) to preserve the text-irrelevant regions, and (3) to merge the manipulated entity into the image naturally. To this end, we propose a new transformer-based framework based on the two-stage image…
The University of Tokyo（东京大学）；University of California Los Angeles（加州大学洛杉矶分校）；AWS AI Labs（AWS人工智能实验室）
Story visualization advances the traditional text-to-image generation by enabling multiple image generation based on a complete story. This task requires machines to 1) understand long text inputs and 2) produce a globally consistent image sequence that illustrates the contents of the story. A key challenge of consistent story visualization is to preserve characters that are essential in stories. To tackle the challenge, we propose to adapt a recent work that augments Vector-Quantized Variational Autoencoders (VQ-VAE) with a text-tovisual-token (transformer) architecture. Specifically, we modi…
University of Electronic Science and Technology of China（电子科技大学）
Detection transformers like DETR have recently shown promising performance on many object detection tasks, but the generalization ability of those methods is still quite challenging for cross-domain adaptation scenarios. To address the cross-domain issue, a straightforward way is to perform token alignment with adversarial training in transformers. However, its performance is often unsatisfactory as the tokens in detection transformers are quite diverse and represent different spatial and semantic information. In this paper, we propose a new method called Spatial-aware and Semantic-aware Token…
Boston University（波士顿大学）；Kuaishou Technology（快手科技）
Most existing methods in vision-language retrieval match two modalities by either comparing their global feature vectors which misses sufficient information and lacks interpretability, detecting objects in images or videos and aligning the text with fine-grained features which relies on complicated model designs, or modeling fine-grained interaction via cross-attention upon visual and textual tokens which suffers from inferior efficiency. To address these limitations, some recent works simply aggregate the token-wise similarities to achieve fine-grained alignment, but they lack intuitive expla…
Beyond Next-Token Alignment: Distilling Multimodal Large Language Models via Token Interactions
MAIS, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所信息与智能系统研究所）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Zhejiang University（浙江大学）
MADTP: Multimodal Alignment-Guided Dynamic Token Pruning for Accelerating Vision-Language Transformer
School of Information Science and Technology, Fudan University（复旦大学信息科学与技术学院）；Academy for Engineering and Technology, Fudan University（复旦大学工程技术研究院）；Tsinghua-Berkeley Shenzhen Institute, Tsinghua University（清华大学深圳国际研究生院（清华-伯克利深圳学院））
Improving CLIP Adaptation by Breaking Tail Alignment for Source-Free Cross-Domain Few-Shot Learning
School of Computer Science and Technology, Huazhong University of Science and Technology, Wuhan, China（华中科技大学计算机科学与技术学院）；Institute of Artificial Intelligence, Huazhong University of Science and Technology, Wuhan, China（华中科技大学人工智能研究院）
Cross-modal Fuzzy Alignment Network for Text-Aerial Person Retrieval and A Large-scale Benchmark
State Key Laboratory of Opto-Electronic Information Acquisition and Protection Technology（光电信息采集与防护技术国家重点实验室）；School of Computer Science and Technology, Anhui University（安徽大学计算机科学与技术学院）；School of Artificial Intelligence, Anhui University（安徽大学人工智能学院）；The University of Hong Kong（香港大学）
Directed-Tokens: A Robust Multi-Modality Alignment Approach to Large Language-Vision Models
CVIU Lab, University of Arkansas, USA（美国亚拉巴马大学CVIU实验室）；Vietnam National University, Ho Chi Minh City University of Science, Vietnam（越南国家大学胡志明市科技大学）；Carnegie Mellon University, USA（美国卡内基梅隆大学）
TokenCLIP: Token-wise Prompt Learning for Zero-shot Anomaly Detection
Zhejiang University（浙江大学）；Tencent YouTu Lab（腾讯YouTu实验室）；Singapore Management University（新加坡管理学院）；Qilu University of Technology (Shandong Academy of Sciences)（青岛科技大学（山东科学院））
Northeastern University（东北大学）；NEC Labs America（NEC美国实验室）；UC San Diego（加州大学圣地亚哥分校）
Self-supervised vision-language pretraining from pure images and text with a contrastive loss is effective, but ignores fine-grained alignment due to a dual-stream architecture that aligns image and text representations only on a global level. Earlier, supervised, non-contrastive methods were capable of finer-grained alignment, but required dense annotations that were not scalable. We propose a single stream architecture that aligns images and language at multiple levels: global, fine-grained patch-token, and conceptual/semantic, using two novel tasks: symmetric cross-modality reconstruction (…
Enhancing Vision-Language Model with Unmasked Token Alignment
Institute for AI Industry Research (AIR)（人工智能产业研究院）；Tsinghua University（清华大学）；Sensetime Research（商汤科技研究院）；CPII under InnoHK（InnoHK下属CPII）；Shanghai AI Laboratory（上海人工智能实验室）
ExpAlign: Expectation-Guided Vision-Language Alignment for Open-Vocabulary Grounding
Department of Automation, Tsinghua University, China（清华大学自动化系）；University of Electronic Science and Technology of China, China（电子科技大学）；University of Copenhagen, Denmark（哥本哈根大学）；Lingsu Lab, China（灵素实验室）
Tsinghua University（清华大学）；Beijing National Research Center for Information Science and Technology（北京信息科学与技术国家研究中心）；Huawei Noah’s Ark Lab（华为诺亚方舟实验室）；Institute for AI Industry Research (AIR), Tsinghua University（清华大学智能产业研究院（AIR））
Many adaptations of transformers have emerged to address the single-modal vision tasks, where self-attention modules are stacked to handle input sources like images. Intuitively, feeding multiple modalities of data to vision transformers could improve the performance, yet the inner-modal attentive weights may also be diluted, which could thus undermine the final performance. In this paper, we propose a multimodal token fusion method (TokenFusion), tailored for transformer-based vision tasks. To effectively fuse multiple modalities, TokenFusion dynamically detects uninformative tokens and subst…
Contrastive learning-based vision-language pre-training approaches, such as CLIP, have demonstrated great success in many vision-language tasks. These methods achieve cross-modal alignment by encoding a matched image-text pair with similar feature embeddings, which are generated by aggregating information from visual patches and language tokens. However, direct aligning cross-modal information using such representations is challenging, as visual patches and text tokens differ in semantic levels and granularities. To alleviate this issue, we propose a Finite Discrete Tokens (FDT) based multimod…
Beijing National Research Center for Information Science and Technology（北京信息科学与技术国家研究中心）；Tsinghua University（清华大学）；PhiGent Robotics（鉴智机器人）
Data mixing strategies (e.g., CutMix) have shown the ability to greatly improve the performance of convolutional neural networks (CNNs). They mix two images as inputs for training and assign them with a mixed label with the same ratio. While they are shown effective for vision transformers (ViTs), we identify a token fluctuation phenomenon that has suppressed the potential of data mixing strategies. We empirically observe that the contributions of input tokens fluctuate as forward propagating, which might induce a different mixing ratio in the output tokens. The training target computed by the…
T2VParser: Adaptive Decomposition Tokens for Partial Alignment in Text to Video Retrieval
Institute of Information Engineering, Chinese Academy of Sciences（中国科学院信息工程研究所）；School of Cyber Security, University of Chinese Academy of Sciences（中国科学院大学网络安全学院）
University of Science and Technology of China（中国科学技术大学）；Microsoft Research Asia（微软亚洲研究院）
Vision transformer has demonstrated great potential in abundant vision tasks. However, it also inevitably suffers from poor generalization capability when the distribution shift occurs in testing (i.e., out-of-distribution data). To mitigate this issue, we propose a novel method, Semantic-aware Message Broadcasting (SAMB), which enables more informative and flexible feature alignment for unsupervised domain adaptation (UDA). Particularly, we study the attention module in the vision transformer and notice that the alignment space using one global class token lacks enough flexibility, where it i…
Microsoft Research（微软研究院）；Carnegie Mellon University（卡内基梅隆大学）
Contrastive learning has been widely used to train transformer-based vision-language models for video-text alignment and multi-modal representation learning. This paper presents a new algorithm called Token-Aware Cascade contrastive learning (TACo) that improves contrastive learning using two novel techniques. The first is the token-aware contrastive loss which is computed by taking into account the syntactic classes of words. This is motivated by the observation that for a video-text pair, the content words in the text, such as nouns and verbs, are more likely to be aligned with the visual co…
Seeing the Image: Prioritizing Visual Correlation by Contrastive Alignment
School of Computer Science, Wuhan University（武汉大学计算机学院）；ByteDance Inc.（字节跳动公司）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）
Decomposed Vision-Language Alignment for Fine-Grained Open-Vocabulary Segmentation
Aerospace Information Research Institute, Chinese Academy of Sciences（中国科学院航空航天信息研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Zhejiang University（浙江大学）
School of Engineering and Built Environment, Electrical and Electronic Engineering, Griffith University（工程与建筑环境学院，电气与电子工程学院，格里菲斯大学）；School of Computer Science and Engineering, University of New South Wales（计算机科学与工程学院，新南威尔士大学）
The University of Hong Kong（香港大学）；University of California, Santa Cruz（加州大学圣克鲁兹分校）；University of Cambridge（剑桥大学）
Learning medical visual representations directly from paired radiology reports has become an emerging topic in representation learning. However, existing medical image-text joint learning methods are limited by instance or local supervision analysis, ignoring disease-level semantic correspondences. In this paper, we present a novel Multi-Granularity Cross-modal Alignment (MGCA) framework for generalized medical visual representation learning by harnessing the naturally exhibited semantic correspondences between medical image and radiology reports at three different levels, i.e., pathological r…
