54 citations · 77 across the 24 of their papers we have counts for
25 papers
Accelerating Visual On-Policy Distillation with Batched Speculative Jacobi Rollouts
Bingqi Shan, Zhehao Yu, Kenhong Lin +1
Visual on-policy distillation (OPD) improves the training of compact visual autoregressive models by learning from trajectories generated by the current student. However, these onl…
Head-Aware Key-Value Compression for Efficient Autoregressive Image Generation
Guotao Liang, Baoquan Zhang, Zhiyuan Wen +1
Autoregressive (AR) visual generation has achieved remarkable performance but suffers from high memory usage and low throughput, as it requires caching previously generated visual…
S2FT: Parameter-Efficient Fine-Tuning in Sparse Spectrum Domain
Baoquan Zhang, Zhehao Yu, Lisai Zhang +5
Parameter Efficient Fine-Tuning (PEFT) is a key technique for adapting a large pretrained model to downstream tasks by fine-tuning only a small number of parameters. Recent methods…
SJD-VP: Speculative Jacobi Decoding with Verification Prediction for Autoregressive Image Generation
Bingqi Shan, Baoquan Zhang, Xiaochen Qi +3
Speculative Jacobi Decoding (SJD) has emerged as a promising method for accelerating autoregressive image generation. Despite its potential, existing SJD approaches often suffer fr…
SJD-PV: Speculative Jacobi Decoding with Phrase Verification for Autoregressive Image Generation
Zhehao Yu, Baoquan Zhang, Bingqi Shan +5
Autoregressive (AR) image models have recently demonstrated remarkable generative capability, but their sequential nature results in significant inference latency. Existing trainin…
Improved Masked Image Generation with Knowledge-Augmented Token Representations
Guotao Liang, Baoquan Zhang, Zhiyuan Wen +2
Masked image generation (MIG) has demonstrated remarkable efficiency and high-fidelity images by enabling parallel token prediction. Existing methods typically rely solely on the m…
GenFT: A Generative Parameter-Efficient Fine-Tuning Method for Pretrained Foundation Models
Guangning Xu, Baoquan Zhang, Michael. K. Ng
Parameter-efficient fine-tuning (PEFT) has emerged as a resource-efficient strategy for adapting Pretrained Foundation Models (PFMs) by learning a small number of task-specific upd…
Towards Improved Text-Aligned Codebook Learning: Multi-Hierarchical Codebook-Text Alignment with Long Text
Guotao Liang, Baoquan Zhang, Zhiyuan Wen +4
Image quantization is a crucial technique in image generation, aimed at learning a codebook that encodes an image into a discrete token sequence. Recent advancements have seen rese…
DeepServe: Serverless Large Language Model Serving at Scale
Junhao Hu, Jiang Xu, Zhixia Liu +18
In this paper, we propose DEEPSERVE, a scalable and serverless AI platform designed to efficiently serve large language models (LLMs) at scale in cloud environments. DEEPSERVE addr…
AsyncDSB: Schedule-Asynchronous Diffusion Schrödinger Bridge for Image Inpainting
Zihao Han, Baoquan Zhang, Lisai Zhang +6
Image inpainting is an important image generation task, which aims to restore corrupted image from partial visible area. Recently, diffusion Schrödinger bridge methods effectively…
Prototype Optimization with Neural ODE for Few-Shot Learning
Baoquan Zhang, Shanshan Feng, Bingqi Shan +3
Few-Shot Learning (FSL) is a challenging task, which aims to recognize novel classes with few examples. Pre-training based methods effectively tackle the problem by pre-training a…
LG-VQ: Language-Guided Codebook Learning
Guotao Liang, Baoquan Zhang, Yaowei Wang +6
Vector quantization (VQ) is a key technique in high-resolution and high-fidelity image synthesis, which aims to learn a codebook to encode an image with a sequence of discrete code…
MCSDNet: Mesoscale Convective System Detection Network via Multi-scale Spatiotemporal Information
Jiajun Liang, Baoquan Zhang, Yunming Ye +3
The accurate detection of Mesoscale Convective Systems (MCS) is crucial for meteorological monitoring due to their potential to cause significant destruction through severe weather…
Codebook Transfer with Part-of-Speech for Vector-Quantized Image Modeling
Baoquan Zhang, Huaibin Wang, Luo Chuyao +5
Vector-Quantized Image Modeling (VQIM) is a fundamental research problem in image synthesis, which aims to represent an image with a discrete token sequence. Existing studies effec…
DiffCast: A Unified Framework via Residual Diffusion for Precipitation Nowcasting
Demin Yu, Xutao Li, Yunming Ye +5
Precipitation nowcasting is an important spatio-temporal prediction task to predict the radar echoes sequences based on current observations, which can serve both meteorological sc…
UER: A Heuristic Bias Addressing Approach for Online Continual Learning
Huiwei Lin, Shanshan Feng, Baoquan Zhang +3
Online continual learning aims to continuously train neural networks from a continuous data stream with a single pass-through data. As the most effective approach, the rehearsal-ba…
HPCR: Holistic Proxy-based Contrastive Replay for Online Continual Learning
Huiwei Lin, Shanshan Feng, Baoquan Zhang +2
Online continual learning, aimed at developing a neural network that continuously learns new data from a single pass over an online data stream, generally suffers from catastrophic…
MetaDiff: Meta-Learning with Conditional Diffusion for Few-Shot Learning
Baoquan Zhang, Chuyao Luo, Demin Yu +4
Equipping a deep model the abaility of few-shot learning, i.e., learning quickly from only few examples, is a core challenge for artificial intelligence. Gradient-based meta-learni…
PCR: Proxy-based Contrastive Replay for Online Class-Incremental Continual Learning
Huiwei Lin, Baoquan Zhang, Shanshan Feng +2
Online class-incremental continual learning is a specific task of continual learning. It aims to continuously learn new classes from data stream and the samples of data stream are…
MetaDT: Meta Decision Tree with Class Hierarchy for Interpretable Few-Shot Learning
Baoquan Zhang, Hao Jiang, Xutao Li +3
Few-Shot Learning (FSL) is a challenging task, which aims to recognize novel classes with few examples. Recently, lots of methods have been proposed from the perspective of meta-le…
SGMNet: Scene Graph Matching Network for Few-Shot Remote Sensing Scene Classification
Baoquan Zhang, Shanshan Feng, Xutao Li +2
Few-Shot Remote Sensing Scene Classification (FSRSSC) is an important task, which aims to recognize novel scene classes with few examples. Recently, several studies attempt to addr…
Prototype Completion for Few-Shot Learning
Baoquan Zhang, Xutao Li, Yunming Ye +1
Few-shot learning aims to recognize novel classes with few examples. Pre-training based methods effectively tackle the problem by pre-training a feature extractor and then fine-tun…
MetaNODE: Prototype Optimization as a Neural ODE for Few-Shot Learning
Baoquan Zhang, Xutao Li, Shanshan Feng +2
Few-Shot Learning (FSL) is a challenging task, \emph{i.e.}, how to recognize novel classes with few examples? Pre-training based methods effectively tackle the problem by pre-train…
Prototype Completion with Primitive Knowledge for Few-Shot Learning
Baoquan Zhang, Xutao Li, Yunming Ye +2
Few-shot learning is a challenging task, which aims to learn a classifier for novel classes with few examples. Pre-training based meta-learning methods effectively tackle the probl…