Xin Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xin Wang has a total of 34 co-authors. Here are the top co-authors by number of shared publications:
- Zhongwei Wan
The Ohio State University, USA
3 shared publications - Mi Zhang
The Ohio State University
2 shared publications - Arvin Hekmati
Viterbi School of Engineering, University of Southern California, Los Angeles, USA
1 shared publication - Bhaskar Krishnamachari
Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, USA
1 shared publication - Boyuan Zheng
University of Michigan, Ann Arbor, USA
1 shared publication - Hangrui Cao
Carnegie Mellon University, USA
1 shared publication
Latest Publications
- Open Access
- 20 April 2026
- Published by MDPI AG in Preprints.org
Speculative Decoding for Multimodal Models: A Survey
Multimodal generative models have demonstrated remarkable capabilities across diverse domains, from visual understanding and image generation to video processing, audio synthesis, and embodied control. These capabilities, however, incur substantial inference overhead due to autoregressive decoding or iterative generation, which is further compounded by modality-specific challenges such as extensive visual token redundancy, strict real-time latency constraints in robotic control, and prolonged sequential generation in text-to-image synthesis. Speculative decoding has emerged as a promising paradigm to accelerate inference without degrading output quality, yet existing surveys remain focused on text-only large language models. In this survey, we provide a systematic and comprehensive review of speculative decoding methods for multimodal models, spanning Vision-Language, Vision-Language-Action, Video-Language, Speech, Text-to-Image (Vision Auto-Regressive), and Diffusion models. We organize the literature into a unified taxonomy with two primary axes, covering the draft generation stage and the verification and acceptance stage, complemented by an analysis of inference framework support. Through this taxonomy, we identify recurring cross-modal design patterns, including token compression, KV cache optimization, target-informed transfer, drafter-target alignment, verification cost reduction, relaxed acceptance, and verify-to-draft feedback, and examine how successful techniques transfer across modalities. We further provide a systematic comparison of existing methods under both self-reported and standardized benchmarking settings. Finally, we discuss open challenges and outline future directions. We have also created a GitHub repository that organizes the papers featured in this survey at https://github.com/zyfzs0/Multimodal-Models-Speculative-Decoding-Survey, and we will actively maintain it as new research emerges. We hope this survey can serve as a valuable resource for researchers and practitioners working on accelerating multimodal inference.
- 20 April 2026
- Published by MDPI AG in Preprints.org
Open AccessSpeculative Decoding for Multimodal Models: A Survey
Multimodal generative models have demonstrated remarkable capabilities across diverse domains, from visual understanding and image generation to video processing, audio synthesis, and embodied control. These capabilities, however, incur substantial inference overhead due to autoregressive decoding or iterative generation, which is further compounded by modality-specific challenges such as extensive visual token redundancy, strict real-time latency constraints in robotic control, and prolonged sequential generation in text-to-image synthesis. Speculative decoding has emerged as a promising paradigm to accelerate inference without degrading output quality, yet existing surveys remain focused on text-only large language models. In this survey, we provide a systematic and comprehensive review of speculative decoding methods for multimodal models, spanning Vision-Language, Vision-Language-Action, Video-Language, Speech, Text-to-Image (Vision Auto-Regressive), and Diffusion models. We organize the literature into a unified taxonomy with two primary axes, covering the draft generation stage and the verification and acceptance stage, complemented by an analysis of inference framework support. Through this taxonomy, we identify recurring cross-modal design patterns, including token compression, KV cache optimization, target-informed transfer, drafter-target alignment, verification cost reduction, relaxed acceptance, and verify-to-draft feedback, and examine how successful techniques transfer across modalities. We further provide a systematic comparison of existing methods under both self-reported and standardized benchmarking settings. Finally, we discuss open challenges and outline future directions. We have also created a GitHub repository that organizes the papers featured in this survey at https://github.com/zyfzs0/Multimodal-Models-Speculative-Decoding-Survey, and we will actively maintain it as new research emerges. We hope this survey can serve as a valuable resource for researchers and practitioners working on accelerating multimodal inference.
- 12 May 2025
- Published by Springer Nature
- p. 268-278
Famba-V: Fast Vision Mamba with Cross-Layer Token Fusion
- 12 May 2025
- Published by Springer Nature
- p. 268-278
Famba-V: Fast Vision Mamba with Cross-Layer Token Fusion
- 16 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet Computing
- Vol. 28 (5) , 57-64
The Internet of Things in the Era of Generative AI: Vision and Challenges
Advancements in Generative AI hold immense promise to push Internet of Things (IoT) to the next level. In this article, we share our vision on IoT in the era of Generative AI. We discuss some of the most important applications of Generative AI in IoT-related domains. We also identify some of the most critical challenges and discuss current gaps as well as promising opportunities on enabling Generative AI for IoT. We hope this article can inspire new research on IoT in the era of Generative AI.
- 16 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet Computing
- Vol. 28 (5) , 57-64
The Internet of Things in the Era of Generative AI: Vision and Challenges
Advancements in Generative AI hold immense promise to push Internet of Things (IoT) to the next level. In this article, we share our vision on IoT in the era of Generative AI. We discuss some of the most important applications of Generative AI in IoT-related domains. We also identify some of the most critical challenges and discuss current gaps as well as promising opportunities on enabling Generative AI for IoT. We hope this article can inspire new research on IoT in the era of Generative AI.
- Open Access
- 13 June 2023
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Management of Data
- Vol. 1 (2) , 1-26
Data Stream Clustering: An In-depth Empirical Study
Data Stream Clustering (DSC) plays an important role in mining continuous and unlabeled data streams in real-world applications. Over the last decades, numerous DSC algorithms have been proposed with promising clustering accuracy and efficiency. Despite the significant differences among existing DSC algorithms, they are commonly built around four key design aspects: summarizing data structure, window model, outlier detection mechanism, and offline refinement strategy. However, there is a lack of empirical studies on these key design aspects in the same codebase using real-world workloads with distinct characteristics. As a result, it is difficult for researchers to improve upon the state-of-the-art. In this paper, we conduct such a study of DSC on its four key design aspects. We implemented state-of-the-art variants of all of these design choices in an open-sourced platform from scratch and evaluated them using both real-world and synthetic workloads. Our analysis identifies the fundamental issues and trade-offs of each design choice in terms of both accuracy and efficiency. We even find that combining flexible design choices led to the development of a new algorithm called Benne, which can be tuned to achieve either better accuracy or better efficiency compared to the state-of-the-art.
- 13 June 2023
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Management of Data
- Vol. 1 (2) , 1-26
Open AccessData Stream Clustering: An In-depth Empirical Study
Data Stream Clustering (DSC) plays an important role in mining continuous and unlabeled data streams in real-world applications. Over the last decades, numerous DSC algorithms have been proposed with promising clustering accuracy and efficiency. Despite the significant differences among existing DSC algorithms, they are commonly built around four key design aspects: summarizing data structure, window model, outlier detection mechanism, and offline refinement strategy. However, there is a lack of empirical studies on these key design aspects in the same codebase using real-world workloads with distinct characteristics. As a result, it is difficult for researchers to improve upon the state-of-the-art. In this paper, we conduct such a study of DSC on its four key design aspects. We implemented state-of-the-art variants of all of these design choices in an open-sourced platform from scratch and evaluated them using both real-world and synthetic workloads. Our analysis identifies the fundamental issues and trade-offs of each design choice in terms of both accuracy and efficiency. We even find that combining flexible design choices led to the development of a new algorithm called Benne, which can be tuned to achieve either better accuracy or better efficiency compared to the state-of-the-art.