Unlocking the Potential of AI

About MediaTek Research

MediaTek Research is a specialized AI research unit within the Global MediaTek Group. With two state- of-the-art research centers located in Cambridge (UK) and National Taiwan University, we foster a collaborative environment where we work closely with esteemed institutions and academics worldwide.

Our team comprises accomplished researchers with diverse backgrounds in computer science, engineering, mathematics, and physics. This expertise enables us to approach the most pressing challenges from multiple angles, fostering innovation and cross-disciplinary collaboration to seek both fundamental breakthroughs and practical applications.

About image 1
About image 2
About image 3
About image 4
About image 5
About image 6
About image 7
About image 8
Vision banner

Vision

Our vision is to push the limits of what is possible in Artificial Intelligence (AI) and Machine Learning (ML). We are committed to advancing the field by developing innovative technologies that empower people, while striving to create systems that are genuinely intelligent, ethical, secure, and sustainable.

Our goal is to enable machines to learn, reason, and interact with humans in ways that are natural, intuitive, and beneficial to society; it should enhance the human potential, enabling us to lead happier, healthier, and more fulfilling lives. We believe that by pushing the boundaries of what AI can do, we can unlock new opportunities, discoveries, and progress that will shape our future.

Research

papers

Papers

Plus icon
Best AI Research Chipset

Tek Talk

Plus icon
MediaTek Research Corp

Seminars

Plus icon
papers

Papers

Plus icon
Best AI Research Chipset

Tek Talk

Plus icon
MediaTek Research Corp

Seminars

Plus icon

Latest Updates

Card
2024-03-05

Breeze-7B: Experience the Latest Highly Efficient Large Language Model Developed by MediaTek Research

MediaTek Advanced Research Center
2023-08-22

Wireless channel modelling with diffusion models at GLOBECOM

MediaTek Research Corp
2023-06-23

Innovate UK awards MediaTek Research up to £1 million funding on Eureka Globalstars.

MediaTek Research
2023-06-23

Improving generative modelling with Shortest Path Diffusion (ICML paper)

MediaTek AI Processing Unit
2023-04-28

MediaTek Research launches the world’s first AI LLM in Traditional Chinese

Best AI Model
2023-04-28

MediaTek Research: Improving the speed and reliability of AI model training

MediaTek Research
2022-10-25

AI breaks into IC design! Deep learning algorithm is showing its power

AI Processing Unit
2022-05-05

MediaTek Announces Breakthrough in Artificial Intelligence and Chip Design

AI Research Chipset
2021-11-16

MediaTek Research opened a brand new office in National Taiwan University

111
2021-11-12

MediaTek Research has 6 papers accepted at NeurIPS 2021 conference and workshops

Field of Expertise

MediaTek AI Processor

Generative
Models

Best AI Processor

Artificial
Intelligence

MediaTek Advanced Research Center

Wireless
Communication

MediaTek AI Processor

Chip
Placement

MediaTek AI Processor

Generative
Models

Best AI Processor

Artificial
Intelligence

MediaTek Advanced Research Center

Wireless
Communication

MediaTek AI Processor

Chip
Placement

Online Lectures

Chang Wei Yueh

Chang Wei Yueh

Trend in AI Theory Seminar: A Theoretical Analysis of Deep Q-Learning

Mark Chang

Mark Chang

Trend in AI Theory Seminar: Provably Efficient Reinforcement Learning Algorithms

Jezabel Garcia

Jezabel Rodriguez Garcia

Trend in AI Theory Seminar: Emergence: Complexity matters also in AI

Yu Wang

Yu Wang

Provable Defenses against Adversarial Examples via the Convex Outer Adversarial Polytope

Yen Ru Lai

Yen Ru Lai

Trends in AI Theory Seminar: Simple And Scalable Off-Policy Reinforcement Learning

Sattar Vakili

Sattar Vakili

An Overview of Stochastic Bandits

Chung En Tsai

Chung En Tsai

Trends in AI Theory Seminar: Learning Quantum States with the Log-Loss

Chiatse Wang

Chiatse Wang

Trends in AI Theory Seminar: An Introduction to Sampling High Dimensional Constrained Continuous..

Sattar Vakili

Sattar Vakili

Improved Convergence Rates for Sparse Approximation Methods in Kernel-Based Learning

Alexandru Cioba

Alexandru Cioba

An Introduction to the Mean-Field Approach for Neural Networks

Jun-Kai You

Jun-Kai You

Polyak-type step sizes for mirror descent methods

Mark Chang

Mark Chang

Optimal Order Simple Regret for Gaussian Process Bandits

Alexandru Cioba

Alexandru Cioba

An Introduction to the Mean-Field Approach for Neural Networks

Mark Chang

Mark Chang

Generative Flow Networks (GFlowNets)

Chung En Tsai

Chung En Tsai

Trends in AI Theory Seminar: "Online Portfolio Selection and Online Entropic Mirror Descent"

Kuan Jen wang

Kuan Jen wang

A brief introduction to optimization on manifolds

Yen Ru Lai

Yen Ru Lai

Off-Policy Deep Reinforcement Learning without Exploration

Mark Chang

Mark Chang

Contextual Bandits with Linear Payoff Functions

Sattar Vakili

Sattar Vakili

Kernel-Based Bandits: Fundamentals and Recent Advances

Jezabel Garcia

Jazabel Rodriguez Garcia

Neural Networks and Quantum Field Theory?

Si-An Chen

Si-An Chen

A Unified View of cGANs with and without Classifiers

Jia-Hau Bau

Jia-Hau Bau

Provable verification on maxpool-based CNN via convex outer bound

Michael Bromberg

Michael Bromberg

Overparametrized Neural Networks and Corresponding Error Estimates

Mark Chang

Mark Chang

Is Q-learning provably efficient?

Alexandru Cioba

Alexandru Cioba

A Brief Recap of SGD Convergence and an Application to MAML

Chang Wei Yueh

Chang Wei Yueh

Trends in AI Theory Seminar: Stochastic bandits robust to adversarial corruptions

Mark Chang

Mark Chang

Trends in AI Theory Seminar: "Contextual Bandits with Linear Payoff Functions"

Alexandru Cioba

Alexandru Cioba

An Introduction to the Mean-Field Approach for Neural Networks

Alexandru Cioba

Alexandru Cioba

A Brief Recap of SGD Convergence and an Application to MAML

Mark Chang

Mark Chang

Contextual Bandits with Linear Payoff Functions

You might also be interested in

MediaTek Al Processing Unit (APU)

MediaTe develops its own Deep Learning Accelerators (Performance Cores), Visual Processing Units (Flexible Cores), hardware-based, multicore scheduler, and software development kits (NeuroPilot) that make up the core components of its industry-leading Al Processing Units (APUs).

MediaTek NeuroPilot

We're meeting the Edge Al challenge head-on with MediaTek NeuroPilot. Through the heterogeneous computing capabilities in our So's such as APUs, GPUs and CPUs, we are providing high-performance and power efficiency for Al features and applications. Developers can target these specific processing units within the chip, or, they can let MediaTe NeuroPilot SD intelligently handle the processing allocation for them.

Contact us background Mobile loading=

Interested in knowing more about
MediaTek Research?

Please feel free to contact us:
info@mtkresearch.com