Publications (28)
You Think, You ACT: The New Task of Arbitrary Text to Motion Generation
Runqi Wang, Caoyuan Ma, Guopeng Li +3
Text to Motion aims to generate human motions from texts. Existing settings rely on limited Action Texts that include action labels, which limits flexibility and practicability in…
FireRed-Image-Edit-1.0 Technical Report
Super Intelligence Team, Changhao Qiao, Chao Hui +16
We present FireRed-Image-Edit, a diffusion transformer for instruction-based image editing that achieves state-of-the-art performance through systematic optimization of data curati…
Cross-Level Distillation and Feature Denoising for Cross-Domain Few-Shot Classification
Hao Zheng, Runqi Wang, Jianzhuang Liu +1
The conventional few-shot classification aims at learning a model on a large labeled base dataset and rapidly adapting to a target dataset that is from the same distribution as the…
Self-Enhancement Improves Text-Image Retrieval in Foundation Visual-Language Models
Yuguang Yang, Yiming Wang, Shupeng Geng +4
The emergence of cross-modal foundation models has introduced numerous approaches grounded in text-image retrieval. However, on some domain-specific retrieval tasks, these models f…
Cogradient Descent for Dependable Learning
Runqi Wang, Baochang Zhang, Li'an Zhuo +2
Conventional gradient descent methods compute the gradients for multiple variables through the partial derivative. Treating the coupled variables independently while ignoring the i…
Anti-Retroactive Interference for Lifelong Learning
Runqi Wang, Yuxiang Bao, Baochang Zhang +3
Humans can continuously learn new knowledge. However, machine learning models suffer from drastic dropping in performance on previous tasks after learning new tasks. Cognitive scie…
VLMInferSlow: Evaluating the Efficiency Robustness of Large Vision-Language Models as a Service
Xiasi Wang, Tianliang Yao, Simin Chen +5
Vision-Language Models (VLMs) have demonstrated great potential in real-world applications. While existing research primarily focuses on improving their accuracy, the efficiency re…
WaveMamba: Wavelet-Driven Mamba Fusion for RGB-Infrared Object Detection
Haodong Zhu, Wenhao Dong, Linlin Yang +10
Leveraging the complementary characteristics of visible (RGB) and infrared (IR) imagery offers significant potential for improving object detection. In this paper, we propose WaveM…
Leader and Follower: Interactive Motion Generation under Trajectory Constraints
Runqi Wang, Caoyuan Ma, Jian Zhao +6
With the rapid advancement of game and film production, generating interactive motion from texts has garnered significant attention due to its potential to revolutionize content cr…
Flux-Sculptor: Text-Driven Rich-Attribute Portrait Editing through Decomposed Spatial Flow Control
Tianyao He, Runqi Wang, Yang Chen +4
Text-driven portrait editing holds significant potential for various applications but also presents considerable challenges. An ideal text-driven portrait editing approach should a…
IdGlow: Dynamic Identity Modulation for Multi-Subject Generation
Honghao Cai, Xiangyuan Wang, Jing Li +15
Multi-subject image generation requires seamlessly harmonizing multiple reference identities within a coherent scene. However, existing methods relying on rigid spatial masks or lo…
Generating Synthetic Computed Tomography for Radiotherapy: SynthRAD2023 Challenge Report
Evi M. C. Huijben, Maarten L. Terpstra, Arthur Jr. Galapon +56
Radiation therapy plays a crucial role in cancer treatment, necessitating precise delivery of radiation to tumors while sparing healthy tissues over multiple days. Computed tomogra…
Edit Where You Mean: Region-Aware Adapter Injection for Mask-Free Local Image Editing
Honghao Cai, Xiangyuan Wang, Yunhao Bai +8
Large diffusion transformers (DiTs) follow global editing instructions well but consistently leak local edits into unrelated regions, because joint-attention architectures offer no…
Unlearning Offline Stochastic Multi-Armed Bandits
Zichun Ye, Runqi Wang, Xuchuang Wang +3
Machine unlearning aims to unlearn data points from a learned model, offering a principled way to process data-deletion requests and mitigate privacy risks without full retraining.…
Confidence Dimension for Deep Learning based on Hoeffding Inequality and Relative Evaluation
Runqi Wang, Linlin Yang, Baochang Zhang +3
Research on the generalization ability of deep neural networks (DNNs) has recently attracted a great deal of attention. However, due to their complex architectures and large number…
Kling-Omni Technical Report
Kling Team, Jialu Chen, Yuanzheng Ci +65
We present Kling-Omni, a generalist generative framework designed to synthesize high-fidelity videos directly from multimodal visual language inputs. Adopting an end-to-end perspec…
DynamicFace: High-Quality and Consistent Face Swapping for Image and Video using Composable 3D Facial Priors
Runqi Wang, Yang Chen, Sijie Xu +6
Face swapping transfers the identity of a source face to a target face while retaining the attributes like expression, pose, hair, and background of the target face. Advanced face…
Single Trajectory Distillation for Accelerating Image and Video Style Transfer
Sijie Xu, Runqi Wang, Wei Zhu +4
Diffusion-based stylization methods typically denoise from a specific partial noise state for image-to-image and video-to-video tasks. This multi-step diffusion process is computat…
Rethinking the Number of Shots in Robust Model-Agnostic Meta-Learning
Xiaoyue Duan, Guoliang Kang, Runqi Wang +4
Robust Model-Agnostic Meta-Learning (MAML) is usually adopted to train a meta-model which may fast adapt to novel classes with only a few exemplars and meanwhile remain robust to a…
P4Q: Learning to Prompt for Quantization in Visual-language Models
Huixin Sun, Runqi Wang, Yanjing Li +4
Large-scale pre-trained Vision-Language Models (VLMs) have gained prominence in various visual and multimodal tasks, yet the deployment of VLMs on downstream application platforms…
Maximum Redundancy Pruning: A Principle-Driven Layerwise Sparsity Allocation for LLMs
Chang Gao, Kang Zhao, Runqi Wang +2
Large language models (LLMs) have demonstrated impressive capabilities, but their enormous size poses significant challenges for deployment in real-world applications. To address t…
Squeeze10-LLM: Squeezing LLMs' Weights by 10 Times via a Staged Mixed-Precision Quantization Method
Qingcheng Zhu, Yangyang Ren, Linlin Yang +9
Deploying large language models (LLMs) is challenging due to their massive parameters and high computational costs. Ultra low-bit quantization can significantly reduce storage and…
Parameter-Efficient Semantic Augmentation for Enhancing Open-Vocabulary Object Detection
Weihao Cao, Runqi Wang, Xiaoyue Duan +3
Open-vocabulary object detection (OVOD) enables models to detect any object category, including unseen ones. Benefiting from large-scale pre-training, existing OVOD methods achieve…
PROMO: Promptable Outfitting for Efficient High-Fidelity Virtual Try-On
Haohua Chen, Tianze Zhou, Wei Zhu +8
Virtual Try-on (VTON) has become a core capability for online retail, where realistic try-on results provide reliable fit guidance, reduce returns, and benefit both consumers and m…
Associative Adversarial Learning Based on Selective Attack
Runqi Wang, Xiaoyue Duan, Baochang Zhang +4
A human's attention can intuitively adapt to corrupted areas of an image by recalling a similar uncorrupted image they have previously seen. This observation motivates us to improv…
Near-Optimal Regret for Efficient Stochastic Combinatorial Semi-Bandits
Zichun Ye, Runqi Wang, Xutong Liu +1
The combinatorial multi-armed bandit (CMAB) is a cornerstone of sequential decision-making framework, dominated by two algorithmic families: UCB-based and adversarial methods such…
AttriCLIP: A Non-Incremental Learner for Incremental Knowledge Learning
Runqi Wang, Xiaoyue Duan, Guoliang Kang +5
Continual learning aims to enable a model to incrementally learn knowledge from sequentially arrived data. Previous works adopt the conventional classification architecture, which…
Few-Shot Learning with Visual Distribution Calibration and Cross-Modal Distribution Alignment
Runqi Wang, Hao Zheng, Xiaoyue Duan +5
Pre-trained vision-language models have inspired much research on few-shot learning. However, with only a few training images, there exist two crucial problems: (1) the visual feat…