# Kaiming He

Kaiming He is a Chinese computer scientist known for co-creating the residual neural network (ResNet) architecture, a foundational deep learning model. He is an associate professor at MIT and a Distinguished Scientist at Google DeepMind, with a career spanning Microsoft Research and Meta AI.

Kaiming He (Chinese: 何恺明; pinyin: Hé Kǎimíng) is a Chinese computer scientist specializing in computer vision and deep learning. He is best known as one of the creators of the residual neural network (ResNet) architecture, which revolutionized the training of deep neural networks. He is an associate professor at the Massachusetts Institute of Technology and works part-time as a Distinguished Scientist at Google DeepMind.

His research has had a profound impact on the field of artificial intelligence, particularly in image recognition and object detection. He has received multiple best paper awards at top conferences and is recognized for his contributions to foundational AI models.

## Early Life and Education

He attended Guangzhou Zhixin High School in Guangzhou, Guangdong, China. In 2003, he achieved the highest total score in the Guangdong provincial undergraduate admissions exam. He then studied at Tsinghua University, earning a Bachelor of Science degree in 2007. From 2007 to 2011, he pursued doctoral studies in information engineering at the Chinese University of Hong Kong, working at its Multimedia Laboratory. He received his PhD in 2011, with a dissertation titled *Single image haze removal using dark channel prior*, advised by Tang Xiao'ou.

## Career and Research

He worked at Microsoft Research Asia from 2011 to 2016, where he contributed to early advances in computer vision. In 2016, he joined Facebook Artificial Intelligence Research (FAIR), now part of Meta, and remained there until 2024. During his tenure at FAIR, he co-authored several influential papers, including the 2016 paper *Deep Residual Learning for Image Recognition*, which introduced the ResNet architecture. This paper became the most cited research paper in a five-year period, according to Google Scholar reports in 2020 and 2021. In 2024, he became an associate professor at the Department of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology.

His work on ResNet addressed the degradation problem in deep neural networks by introducing skip connections, enabling the training of networks with hundreds or thousands of layers. This innovation underpins many modern AI systems, including those used in [machine-learning](https://www.wikiprompt.org/wiki/machine-learning) and [artificial-intelligence](https://www.wikiprompt.org/wiki/artificial-intelligence). He also contributed to the development of Mask R-CNN, a framework for instance segmentation, which extends [neural-network](https://www.wikiprompt.org/wiki/neural-network) approaches to pixel-level object detection.

## Awards and Recognitions

He has received several prestigious awards. He won the Marr Prize for the best paper at the International Conference on Computer Vision (ICCV) in 2017. He also won the best paper award at the Computer Vision and Pattern Recognition (CVPR) conference in 2009 and 2016. In 2023, he was awarded the Future Science Prize, along with three collaborators, for their fundamental contributions to artificial intelligence through the introduction of deep residual learning. His work is widely cited in the fields of [deep-learning](https://www.wikiprompt.org/wiki/deep-learning) and computer vision.

## Impact and Legacy

He's research has influenced the development of various AI applications, from image classification to autonomous driving. The ResNet architecture is a standard component in many modern systems, including those used in [google-deepmind](https://www.wikiprompt.org/wiki/google-deepmind) and other leading AI labs. His contributions have also shaped the work of researchers at institutions like [stanford-ai-lab](https://www.wikiprompt.org/wiki/stanford-ai-lab) and [berkeley-ai-research](https://www.wikiprompt.org/wiki/berkeley-ai-research). His approach to model design has been adopted in [transformer](https://www.wikiprompt.org/wiki/transformer) architectures and other advanced models, highlighting the cross-pollination of ideas in the AI community.

## Current Work

As of 2024, he is based at MIT, where he continues to research computer vision and deep learning. His part-time role at Google DeepMind involves collaborating on projects that push the boundaries of AI. His ongoing work focuses on improving the efficiency and capability of neural networks, with implications for [generative-ai](https://www.wikiprompt.org/wiki/generative-ai) and other emerging technologies.

---
Source: https://www.wikiprompt.org/wiki/kaiming-he
License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)
Last updated: 2026-09-05T13:26:16.204185+00:00