Author name cluster

Tao Wei

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

9 papers

2 author rows

AAAI Conference 2026 Conference Paper

SRSplat: Feed-Forward Super-Resolution Gaussian Splatting from Sparse Multi-View Images

Xinyuan Hu
Changyue Shi
Chuxiao Yang
Minghao Chen
Jiajun Ding
Tao Wei
Chen Wei
Zhou Yu

Feed-forward 3D reconstruction from sparse, low-resolution (LR) images is a crucial capability for real-world applications, such as autonomous driving and embodied AI. However, existing methods often fail to recover fine texture details. This limitation stems from the inherent lack of high-frequency information in LR inputs. To address this, we propose SRSplat, a feed-forward framework that reconstructs high-resolution 3D scenes from only a few LR views. Our main insight is to compensate for the deficiency of texture information by jointly leveraging external high-quality reference images and internal texture cues. We first construct a scene-specific reference gallery, generated for each scene using Multimodal Large Language Models (MLLMs) and diffusion models. To integrate this external information, we introduce the Reference-Guided Feature Enhancement (RGFE) module, which aligns and fuses features from the LR input images and their reference twin image. Subsequently, we train a decoder to predict the Gaussian primitives using the multi-view fused feature obtained from RGFE. To further refine predicted Gaussian primitives, we introduce Texture-Aware Density Control (TADC), which adaptively adjusts Gaussian density based on the internal texture richness of the LR inputs. Extensive experiments demonstrate that our SRSplat outperforms existing methods on various datasets, including RealEstate10K, ACID, and DTU, and exhibits strong cross-dataset and cross-resolution generalization capabilities.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

AegisGuard: RL-Guided Adapter Tuning for TEE-Based Efficient & Secure On-Device Inference

CHE WANG
Ziqi Zhang
Yinggui Wang
Tiantong Wang
Yurong Hao
Jianbo Gao
Tao Wei
Yang Cao

On-device large models (LMs) reduce cloud dependency but expose proprietary model weights to the end-user, making them vulnerable to white-box model stealing (MS) attacks. A common defense is TEE-Shielded DNN Partition (TSDP), which places all trainable LoRA adapters (fine tuned on private data) inside a trusted execution environment (TEE). However, this design suffers from excessive host-to-TEE communication latency. We propose AegisGuard, a fine tuning and deployment framework that selectively shields the MS sensitive adapters while offloading the rest to the GPU, balancing security and efficiency. AegisGuard integrates two key components: i) RL-based Sensitivity Measurement (RSM), which injects Gaussian noise during training and applies a lightweight reinforcement learning to rank adapters based on their impact on model stealing; and (ii) Shielded-Adapter Compression (SAC), which structurally prunes the selected adapters to reduce both parameter size and intermediate feature maps, further lowering TEE computation and data transfer costs. Extensive experiments demonstrate that AegisGuard achieves black-box level MS resilience (surrogate accuracy around 39%, matching fully shielded baselines), while reducing end-to-end inference latency by 2–3× and cutting TEE memory usage by 4× compared to state-of-the-art TSDP methods.

PDF Details

ICML Conference 2025 Conference Paper

Kona: An Efficient Privacy-Preservation Framework for KNN Classification by Communication Optimization

Guopeng Lin
Ruisheng Zhou
Shuyu Chen
Weili Han
Jin Tan
Wenjing Fang
Lei Wang
Tao Wei

K-nearest neighbors (KNN) classification plays a significant role in various applications due to its interpretability. The accuracy of KNN classification relies heavily on large amounts of high-quality data, which are often distributed among different parties and contain sensitive information. Dozens of privacy-preserving frameworks have been proposed for performing KNN classification with data from different parties while preserving data privacy. However, existing privacy-preserving frameworks for KNN classification demonstrate communication inefficiency in the online phase due to two main issues: (1) They suffer from huge communication size for secure Euclidean square distance computations. (2) They require numerous communication rounds to select the $k$ nearest neighbors. In this paper, we present $\texttt{Kona}$, an efficient privacy-preserving framework for KNN classification. We resolve the above communication issues by (1) designing novel Euclidean triples, which eliminate the online communication for secure Euclidean square distance computations, (2) proposing a divide-and-conquer bubble protocol, which significantly reduces communication rounds for selecting the $k$ nearest neighbors. Experimental results on eight real-world datasets demonstrate that $\texttt{Kona}$ significantly outperforms the state-of-the-art framework by $1. 1\times \sim 3121. 2\times$ in communication size, $16. 7\times \sim 5783. 2\times$ in communication rounds, and $1. 1\times \sim 232. 6\times$ in runtime.

Details

NeurIPS Conference 2025 Conference Paper

MPCache: MPC-Friendly KV Cache Eviction for Efficient Private LLM Inference

Wenxuan Zeng
Ye Dong
Jinjin Zhou
Jin Tan
Lei Wang
Tao Wei
Runsheng Wang
Meng Li

Private large language model (LLM) inference based on secure multi-party computation (MPC) achieves formal data privacy protection but suffers from significant latency overhead, especially for long input sequences. While key-value (KV) cache eviction and sparse attention algorithms have been proposed for efficient LLM inference in plaintext, they are not designed for MPC and cannot benefit private LLM inference directly. In this paper, we propose an accurate and MPC-friendly KV cache eviction framework, dubbed MPCache, building on the observation that historical tokens in a long sequence may have different effects on the downstream decoding. Hence, MPCache combines a look-once static eviction algorithm to discard unimportant KV cache and a query-aware dynamic selection algorithm to activate only a small subset of KV cache for attention computation. MPCache further incorporates a series of optimizations for efficient dynamic KV cache selection, including MPC-friendly similarity approximation, hierarchical KV cache clustering, and cross-layer index-sharing strategy. Extensive experiments demonstrate that MPCache consistently outperforms prior-art KV cache eviction baselines across different generation tasks and achieves 1. 8 ~ 2. 01x and 3. 39 ~ 8. 37x decoding latency and communication reduction on different sequence lengths, respectively.

PDF Details

JBHI Journal 2024 Journal Article

Boundary-Sensitive Segmentation of Small Liver Lesions

Tao Wei
Yiqi Wang
Yuqiang Zhang
Yunfu Wang
Liang Zhao

Early diagnosis plays a pivotal role in handling the global health challenge posed by liver diseases. However, early-stage lesions are typically quite small, presenting significant difficulties due to insufficient regions for developing effective features, indistinguishable boundaries of small lesions, and a lack of tiny liver lesion masks. To address these issues, we approach the solution in two-fold: an efficient model and a high-quality dataset. The model is built upon the advantages of path signature and camouflaged object detection. The path signature narrows down the ambiguous boundaries between lesions and other tissues while the camouflaged object detection achieves high accuracy in detecting inconspicuous lesions. The two are seamlessly integrated to ensure high accuracy and fidelity. For the dataset, we collect more than ten thousand liver images with over four thousand lesions, approximately half of which are small. Experiments on both an established dataset and our newly constructed one show that the proposed model outperforms state-of-the-art semantic segmentation and camouflaged object detection models, particularly in detecting small lesions. Moreover, the decisive and faithful salience maps generated by the model at the boundary regions demonstrate its strong robustness.

Details DOI

AAAI Conference 2024 Conference Paper

Dual-Prior Augmented Decoding Network for Long Tail Distribution in HOI Detection

Jiayi Gao
Kongming Liang
Tao Wei
Wei Chen
Zhanyu Ma
Jun Guo

Human object interaction detection aims at localizing human-object pairs and recognizing their interactions. Trapped by the long-tailed distribution of the data, existing HOI detection methods often have difficulty recognizing the tail categories. Many approaches try to improve the recognition of HOI tasks by utilizing external knowledge (e.g. pre-trained visual-language models). However, these approaches mainly utilize external knowledge at the HOI combination level and achieve limited improvement in the tail categories. In this paper, we propose a dual-prior augmented decoding network by decomposing the HOI task into two sub-tasks: human-object pair detection and interaction recognition. For each subtask, we leverage external knowledge to enhance the model's ability at a finer granularity. Specifically, we acquire the prior candidates from an external classifier and embed them to assist the subsequent decoding process. Thus, the long-tail problem is mitigated from a coarse-to-fine level with the corresponding external knowledge. Our approach outperforms existing state-of-the-art models in various settings and significantly boosts the performance on the tail HOI categories. The source code is available at https://github.com/PRIS-CV/DP-ADN.

PDF Details DOI

ICML Conference 2016 Conference Paper

Network Morphism

Tao Wei
Changhu Wang
Yong Rui
Chang Wen Chen

We present a systematic study on how to morph a well-trained neural network to a new one so that its network function can be completely preserved. We define this as network morphism in this research. After morphing a parent network, the child network is expected to inherit the knowledge from its parent network and also has the potential to continue growing into a more powerful one with much shortened training time. The first requirement for this network morphism is its ability to handle diverse morphing types of networks, including changes of depth, width, kernel size, and even subnet. To meet this requirement, we first introduce the network morphism equations, and then develop novel morphing algorithms for all these morphing types for both classic and convolutional neural networks. The second requirement is its ability to deal with non-linearity in a network. We propose a family of parametric-activation functions to facilitate the morphing of any continuous non-linear activation neurons. Experimental results on benchmark datasets and typical neural networks demonstrate the effectiveness of the proposed network morphism scheme.

Details

ICRA Conference 2006 Conference Paper

Particle Filtering for Adaptive Sensor Fault Detection and Identification

Tao Wei
Yufei Huang 0001
C. L. Philip Chen

In this paper, we address the problem of adaptive sensor fault identification and validation by particle filtering. The model-based approaches are developed, where the sensor system is modeled by a Markov switch dynamic state-space model. To handle the nonlinearity of the problem, two different particle filters: mixture Kalman filter (MKF) and stochastic M-algorithm (SMA) are proposed. Simulation results are presented to compare the effectiveness and complexity of MKF and SMA methods

Details

ICRA Conference 2001 Conference Paper

Development of a Sheet-Based Material Handling System for Layered Manufacturing

Tao Wei
Sangeun Choi
Wyatt S. Newman

Computer-aided manufacturing of laminated engineering materials (CAM-LEM), developed at Case Western Reserve University, is a layered-manufacturing approach utilizing automated stacking of parts laser-cut from sheet materials. This paper presents the design of a material-handling system and its integration with a modular laser-cutting system, completing the most recent version of CAM-LEM. Our material-handling system is required to: transfer individual sheets from at least three material stacks to the cutting table of the laser-cutting system; extract single-layer, laser-cut parts from the cutting table; and stack the parts onto a vertical-stack assembly, resulting in fabrication of a 3D object. Design philosophy, details, and stacking performance measurements of the constructed system are presented. It is shown that the resulting system is capable of assembling objects within a 150 mm cube with an assembly precision of approximately 0. 050 mm. Further, the assembly operations are relatively fast, so that part build rates are limited primarily by the laser cutter.

Details