Publications (17)
Beyond the Federation: Topology-aware Federated Learning for Generalization to Unseen Clients
Mengmeng Ma, Tang Li, Xi Peng
Federated Learning is widely employed to tackle distributed sensitive data. Existing methods primarily focus on addressing in-federation data heterogeneity. However, we observed th…
Inside the Visual Mind: Neuroscience-Motivated Concept Circuits for Interpreting and Steering Vision Transformers
Tang Li, Yanlin Chen, Mengmeng Ma +1
Despite high accuracy, Vision Transformer (ViT) predictions can be driven by spurious cues, raising the need to understand their inner workings before safe deployment. Sparse autoe…
Beyond Accuracy: Ensuring Correct Predictions With Correct Rationales
Tang Li, Mengmeng Ma, Xi Peng
Large pretrained foundation models demonstrate exceptional performance and, in some high-stakes applications, even surpass human experts. However, most of these models are currentl…
Towards Human-Like Machine Comprehension: Few-Shot Relational Learning in Visually-Rich Documents
Hao Wang, Tang Li, Chenhui Chu +3
Key-value relations are prevalent in Visually-Rich Documents (VRDs), often depicted in distinct spatial regions accompanied by specific color and font styles. These non-textual cue…
Observation of Fundamental Thermal Noise in Optical Fibers down to Infrasonic Frequencies
Jing Dong, Junchao Huang, Tang Li +1
The intrinsic thermal noise in optical fibers is the ultimate limit of fiber-based systems. However, at infrasonic frequencies, the spectral behavior of the intrinsic thermal noise…
Streaming Voice Conversion Via Intermediate Bottleneck Features And Non-streaming Teacher Guidance
Yuanzhe Chen, Ming Tu, Tang Li +7
Streaming voice conversion (VC) is the task of converting the voice of one person to another in real-time. Previous streaming VC methods use phonetic posteriorgrams (PPGs) extracte…
Deep Learning for Spatiotemporal Modeling of Urbanization
Tang Li, Jing Gao, Xi Peng
Urbanization has a strong impact on the health and wellbeing of populations across the world. Predictive spatial modeling of urbanization therefore can be a useful tool for effecti…
Cold Atom Space Clock with Counter-Propagating Atoms
Desheng Lv, Bin Wang, Tang Li +1
We discuss the feasibility to realize a space cold atom clock with counter-propagating cold atoms in microgravity. The design of the space clock is based on atomic beam clock with…
Beyond Accuracy: On the Effects of Fine-tuning Towards Vision-Language Model's Prediction Rationality
Qitong Wang, Tang Li, Kien X. Nguyen +1
Vision-Language Models (VLMs), such as CLIP, have already seen widespread applications. Researchers actively engage in further fine-tuning VLMs in safety-critical domains. In these…
Efficient Neural Music Generation
Max W. Y. Lam, Qiao Tian, Tang Li +10
Recent progress in music generation has been remarkably advanced by the state-of-the-art MusicLM, which comprises a hierarchy of three LMs, respectively, for semantic, coarse acous…
Application-level Studies of Cellular Neural Network-based Hardware Accelerators
Qiuwen Lou, Indranil Palit, Tang Li +3
As cost and performance benefits associated with Moore's Law scaling slow, researchers are studying alternative architectures (e.g., based on analog and/or spiking circuits) and/or…
X-ray multibeam ptychography at up to 20 keV: nano-lithography enhances X-ray nano-imaging
Tang Li, Maik Kahnt, Thomas L. Sheppard +6
Non-destructive nano-imaging of the internal structure of solid matter is only feasible using hard X-rays due to their high penetration. The highest resolution images are achieved…
Tests of Cold Atom Clock in Orbit
Liang Liu, Desheng Lü, Weibiao Chen +23
Since the atomic clock was invented, its performance has been improved for one digit every decade until 90s of last century when the traditional atomic clock almost reached its lim…
Interpretable Failure Detection with Human-Level Concepts
Kien X. Nguyen, Tang Li, Xi Peng
Reliable failure detection holds paramount importance in safety-critical applications. Yet, neural networks are known to produce overconfident predictions for misclassified samples…
PolyVoice: Language Models for Speech to Speech Translation
Qianqian Dong, Zhiying Huang, Qiao Tian +15
We propose PolyVoice, a language model-based framework for speech-to-speech translation (S2ST) system. Our framework consists of two language models: a translation language model a…
DEAL: Disentangle and Localize Concept-level Explanations for VLMs
Tang Li, Mengmeng Ma, Xi Peng
Large pre-trained Vision-Language Models (VLMs) have become ubiquitous foundational components of other models and downstream tasks. Although powerful, our empirical results reveal…
Are Data-driven Explanations Robust against Out-of-distribution Data?
Tang Li, Fengchun Qiao, Mengmeng Ma +1
As black-box models increasingly power high-stakes applications, a variety of data-driven explanation methods have been introduced. Meanwhile, machine learning models are constantl…