Efficient First-Order Optimization on the Pareto Set for Multi-Objective Learning under Preference Guidance

no code implementations • 26 Mar 2025 • Lisha Chen, Quan Xiao, Ellen Hidemi Fukuda, Xinyi Chen, Kun Yuan, Tianyi Chen

To solve this problem, we convert the multi-objective constraints to a single-objective constraint through a merit function with an easy-to-evaluate gradient, and then, we use a penalty-based reformulation of the bilevel optimization problem.

Bilevel Optimization Fairness +2

Bilevel Joint Unsupervised and Supervised Training for Automatic Speech Recognition

no code implementations • 11 Dec 2024 • Xiaodong Cui, A F M Saif, Songtao Lu, Lisha Chen, Tianyi Chen, Brian Kingsbury, George Saon

In this paper, we propose a bilevel joint unsupervised and supervised training (BL-JUST) framework for automatic speech recognition.

Automatic Speech Recognition speech-recognition +1

FERERO: A Flexible Framework for Preference-Guided Multi-Objective Learning

1 code implementation • 2 Dec 2024 • Lisha Chen, AFM Saif, Yanning Shen, Tianyi Chen

In this work, we introduce a Flexible framEwork for pREfeRence-guided multi-Objective learning (FERERO) by casting it as a constrained vector optimization problem.

Learning with Limited Samples -- Meta-Learning and Applications to Communication Systems

no code implementations • 3 Oct 2022 • Lisha Chen, Sharu Theresa Jose, Ivana Nikoloska, Sangwoo Park, Tianyi Chen, Osvaldo Simeone

This review monograph provides an introduction to meta-learning by covering principles, algorithms, theory, and engineering applications.

Bilevel Optimization image-classification +4

Understanding Benign Overfitting in Gradient-Based Meta Learning

no code implementations • 27 Jun 2022 • Lisha Chen, Songtao Lu, Tianyi Chen

While the conventional statistical learning theory suggests that overparameterized models tend to overfit, empirical evidence reveals that overparameterized meta learning methods still work well -- a phenomenon often called "benign overfitting."

Few-Shot Learning Learning Theory

Sharp-MAML: Sharpness-Aware Model-Agnostic Meta Learning

1 code implementation • 8 Jun 2022 • Momin Abbas, Quan Xiao, Lisha Chen, Pin-Yu Chen, Tianyi Chen

Model-agnostic meta learning (MAML) is currently one of the dominating approaches for few-shot meta-learning.

Meta-Learning model

Is Bayesian Model-Agnostic Meta Learning Better than Model-Agnostic Meta Learning, Provably?

1 code implementation • 6 Mar 2022 • Lisha Chen, Tianyi Chen

In this paper, we aim to provide theoretical justifications for Bayesian MAML's advantageous performance by comparing the meta test risks of MAML and Bayesian MAML.

Meta-Learning model

Deep Structured Prediction for Facial Landmark Detection

1 code implementation • NeurIPS 2019 • Lisha Chen, Hui Su, Qiang Ji

Existing deep learning based facial landmark detection methods have achieved excellent performance.

Face Alignment Facial Landmark Detection +2

Face Alignment With Kernel Density Deep Neural Network

no code implementations • ICCV 2019 • Lisha Chen, Hui Su, Qiang Ji

Specifically, for face alignment, we adapt state-of-the-art hourglass neural network into a probabilistic neural network framework with landmark probability map as its output.

Face Alignment