Fei-Fei Li

Personal Profile

Fei Fei Li (English name: Fei Fei Li), born in 1976, is a Chinese American computer scientist known for establishing ImageNet, a dataset that propelled the rapid development of computer vision in the 2010s.

She is a Sequoia Capital professor of computer science at Stanford University and a former board member of Twitter. Li Feifei is the co director of the HAI Institute at Stanford and also the co director of the Stanford Vision and Learning Laboratory. From 2013 to 2018, she served as the director of the Stanford Artificial Intelligence Laboratory (SAIL).

In 2017, she co founded AI4ALL, a non-profit organization dedicated to improving diversity and inclusivity in the field of artificial intelligence. Her research expertise includes artificial intelligence (AI), machine learning, deep learning, computer vision, and cognitive neuroscience.

Due to his contributions in building large-scale knowledge bases for machine learning and visual understanding, Li Feifei was elected as a member of the National Academy of Engineering (NAE) in the United States in 2020. She is also a member of the National Academy of Medicine (NAM) and the American Academy of Arts and Sciences (AAAS).

Early and educational experiences

Li Feifei was born in Beijing in 1976 and grew up in Chengdu. At the age of 12, her father moved to the United States. At the age of 15, she and her mother reunited with their father in Parsippany Troy Mountain, New Jersey. She graduated from Parsippani High School in 1995 and was inducted into the Parsippani High School Hall of Fame in 2017.

In 1999, she graduated with honors from Princeton University with a Bachelor’s degree in Physics and certificates in Applied and Computational Mathematics and Engineering Physics. Under the guidance of Professor Bradley Dickinson in Electrical Engineering, she completed her graduation thesis titled “Auditory Binaral Correspondence Difference: A New Computational Model for Huggins Dichoic Pitch”. Subsequently, Li Feifei pursued graduate studies at the California Institute of Technology and obtained a PhD in Electronic Engineering in 2005.

Work Experience

From 2005 to August 2009, Li Feifei served as an assistant professor in the Department of Electronic and Computer Engineering at the University of Illinois at Urbana Champaign and the Department of Computer Science at Princeton University. She joined Stanford University in 2009 as an assistant professor, was promoted to a tenured associate professor in 2012, and became a full-time professor in 2017. At Stanford University, Li Feifei served as the director of the Stanford Artificial Intelligence Laboratory (SAIL) from 2013 to 2018. She became the founding co director of Stanford University’s university level initiative, the Human Centered Artificial Intelligence Institute (HAI), along with former Dean of Academic Affairs Dr. John Echizendi.

During his vacation from Stanford University from January 2017 to the fall of 2018, Li Feifei joined Google Cloud as its Chief Scientist and Vice President of Artificial Intelligence/Machine Learning. At Google, her team focuses on democratizing artificial intelligence technology, lowering entry barriers for businesses and developers, including the development of products such as AutoML. In the autumn of 2018, Li Feifei left Google and returned to Stanford University to continue her professorship.

As co-founder and chairman of the non-profit organization AI4ALL, Li Feifei is also known for his non-profit work. The organization’s mission is to promote diversity and inclusivity through people-centered artificial intelligence principles, and to educate the next generation of AI technology experts, thinkers, and leaders.

Before establishing AI4ALL in 2017, Li Feifei and her former student Olga Russakovsky (currently an assistant professor at Princeton University) co founded and co directed a pioneering project called SAILORS (Stanford AI Lab OutReach Summers) at Stanford University. SAILORS is Stanford’s annual summer camp dedicated to AI education and research for ninth grade high school girls. It was founded in 2015 and was renamed AI4ALL @ Stanford until 2017. In 2018, in addition to Stanford University, AI4ALL successfully launched five summer programs, including Princeton University, Carnegie Mellon University, Boston University, University of California, Berkeley, and Simon Fraser University in Canada.

From ImageNet to Spatial Intelligence.

The “AI Persona” aims to document the key figures who have influenced the historical development of AI, providing a glimpse into the historical inevitability, current focus, and future trends of the AI industry through their choices in education and work.

She is an AI godmother who equips machines with “eyes” and, together with her team, created the ImageNet dataset, driving the wave of deep learning; She led a team to develop Google Cloud AutoML, which lowered the threshold for using AI technology and made it easy for small and medium-sized enterprises to use intelligent tools.

She is still an explorer in the field of space intelligence, dedicating herself to world models that enable AI to understand and predict spatial relationships through her keen insight into the physical world.

She once struggled to survive in the kitchen of a restaurant, but rose to the podium of the world’s top universities with her love for science; She has faced controversy for adhering to the ethical principles of AI, but has never changed her original intention of “technology for good and people-oriented”.

She has proven through decades of exploration that the ultimate significance of AI lies not in cold algorithms, but in its deep concern for human needs.

Today, we embark on the life journey of “AI godmother” Li Feifei, exploring how this “data pioneer” continuously promotes the development of the AI industry through love and persistence.

The AI Dream of Immigrant Girls

Li Feifei was born in Beijing in 1976. His father is an engineer and his mother is a teacher. The engineering drawings and electronic components in the study became her earliest science enlightenment textbooks.

After moving to Sichuan with her family in her childhood, the strong academic atmosphere of the middle school she attended allowed her natural talent in science to fully develop. She likes to dismantle old appliances to study their principles, or use her pocket money to purchase electronic components for experiments. This passion for hands-on exploration planted the seeds for her future involvement in scientific research.

In 1992, 15-year-old Li Feifei went to the United States, and contrary to expectations, the reality of the difficulties far exceeded imagination. My parents lost their decent jobs and the whole family had to squeeze into a small house in the town of Parsippany.

In order to share the financial pressure of his family, former top student Li Feifei decided to put down his books and work tirelessly in the kitchen of restaurants in New York’s Chinatown for more than ten hours every day. The greasy tableware, noisy environment, and unfamiliar language did not diminish her thirst for knowledge. After closing, the dining table became a desk, the dictionary was curled up, the TV news became an English textbook, and the four o’clock light in the morning witnessed her persistence

Finally, with the rapid improvement of her grades, she tore off the label of “immigrant underachiever” in high school and achieved a perfect score of 1250 in mathematics during the college entrance examination.

After graduating from high school, Li Feifei entered the Physics Department of Princeton University and received a full scholarship. The rigorous thinking training in physics laid a solid logical foundation for her later AI research.

In order to earn enough living expenses, Li Feifei’s family borrowed money to open a dry cleaning shop and began a busy life of “studying for five days and working on weekends”.

ImageNet that changes the world

In 1999, Li Feifei graduated from Princeton University with excellent grades. At this moment, she is facing another important decision in her life. With the prestigious reputation of Princeton University, she has received job offers from several Wall Street financial giants, including Goldman Sachs. However, Li Feifei made an unexpected decision to give up the high paying job opportunity and choose to go to Xizang to study Tibetan medicine.

For Li Feifei, studying Tibetan medicine is not a momentary impulse. She has always had a profound understanding and attention to the significance of niche scientific research projects in a larger field. In her opinion, Tibetan medicine can bring her more inspiration and thinking at the philosophical and methodological levels. During her stay in Xizang, she deeply studied the pharmacology and efficacy of Tibetan medicine, exchanged and learned with local Tibetan medicine, and experienced the profound Tibetan medicine culture.

After returning from Xizang in 2002, Li Feifei decided to enter the California Institute of Technology to study for a doctorate in AI and computational neuroscience. At that time, the field of computer vision was still in its infancy, but the types of objects that computers could recognize were very limited, and many theories and technologies still needed to be explored and improved.

But Li Feifei firmly believes that computer vision recognition has broad application prospects and is of great significance in promoting the development of AI. So she resolutely chose this path full of thorns. Moreover, during her doctoral studies, her mother was diagnosed with cancer, which dealt a heavy blow to her life and studies. With her tenacious perseverance and persistent pursuit of scientific research, she took care of her sick mother while working hard to complete her studies.

At the beginning, Li Feifei devoted a lot of energy to algorithm optimization. She leads the team to improve and innovate existing algorithms. However, they found that relying solely on algorithm optimization, the accuracy of computer vision recognition still cannot meet the needs of practical applications. After experiencing multiple failures, Li Feifei began to reflect on his research approach. She gradually realized that the key to teaching computers to recognize images was to enable them to see more images, which required rich data support.

So, Li Feifei decided to embark on an unprecedented project – to establish a massive image database. She plans to download a large number of pictures from the Internet, classify and label these pictures, and provide the computer with a “question bank” for learning. This project is ImageNet, which later promoted the development of the AI industry.

In 2006, Li Feifei returned to Princeton University and devoted himself fully to the ImageNet project. Her goal was to establish an image dataset containing up to 30000 categories, which was an extremely bold and challenging idea at the time. Many people are skeptical of her project, believing it to be an almost impossible task to complete. However, Li Feifei was not shaken by external doubts and firmly believed that her direction was correct.

At the beginning of the project, Li Feifei encountered many difficulties. The first is the problem of data collection. It is not easy to download a large number of pictures from the Internet, which requires a lot of time and energy, and also faces legal issues such as copyright. Secondly, there is the challenge of data annotation. If manual annotation is used, it not only requires a lot of manpower and financial resources, but also takes a long time. According to estimates at the time, it would take 19 years to annotate a dataset of 30000 types of images.

Fortunately, Li Feifei met two important supporters. One is Professor Kai Li from the Department of Computer Science at Princeton University, who believes that Li Feifei’s research direction has great potential. He not only gave her a set of workstations, but also “transferred” his graduate student Deng Jia to her to assist in carrying out research work. The other is Sun Min, who introduced Amazon’s “Türkiye Robot” crowdsourcing platform to Li Feifei. Through this platform, Li Feifei can distribute image annotation work to people around the world, greatly improving the efficiency of annotation and reducing costs.

By 2009, the ImageNet database had contained 15 million annotated images, unprecedented in both quality and quantity in the scientific community. More importantly, Li Feifei has made ImageNet, a massive image database, available for free use. This initiative has milestone significance, as it means that all teams dedicated to computer vision recognition worldwide can access data and test questions from this database to train and test the accuracy of their own algorithms.

The emergence of ImageNet has propelled the rapid development of the entire field of computer vision.

Li Feifei, Stanford HAI

In 2009, Li Feifei joined Stanford University as an assistant professor. Here, she continues to delve deeper into computer vision research. She led a team to design an algorithm that pairs convolutional neural network technology with recursive neural networks in natural language processing technology. This not only allows machines to annotate objects appearing in front of them, but also enables them to describe the entire scene. This is a groundbreaking technological advancement that opens up new avenues for the application of artificial intelligence in image understanding and description.

In 2012, Li Feifei ushered in another important moment in her academic career – being awarded the title of Lifetime Associate Professor at Stanford University. From 2013 to 2018, she served as the director of the Artificial Intelligence Laboratory at Stanford University. Under her leadership, the laboratory achieved multiple important research results in the field of artificial intelligence and became one of the important bases for global AI research.

At the end of 2016, Li Feifei made a surprising decision: to temporarily leave Stanford University and become the Chief Scientist of Google Cloud. If the technology in the laboratory cannot be applied in practice, it will only be beautiful papers. Her goal is very clear, which is to promote the popularization of AI. At that time, AI technology was mainly controlled by a few tech giants, making it difficult for small and medium-sized enterprises to access and apply it. The Google Cloud AutoML platform she led the development of completely changed this situation.

This automated tool allows non professional users to train AI models: employees of flower planting companies can upload pictures and simply annotate them to obtain accurate flower recognition systems; Farmers can quickly diagnose pests and diseases by taking crop photos with their mobile phones. After the platform was launched, the registered users exceeded one million within a few months. Small restaurants used it to optimize their ordering systems, and museums relied on it to digitize cultural relics. Li Feifei’s philosophy has truly brought AI into people’s daily lives.

In the autumn of 2018, under the promotion of Li Feifei, the Stanford Human Centered Artificial Intelligence Institute (Stanford HAI) began construction and was officially established in 2019. The establishment of HAI aims to promote the development of AI technology and better serve human society. HAI aims to gather top AI experts and scholars from around the world to conduct interdisciplinary research and explore how AI technology can bring convenience to humanity while avoiding potential negative impacts.

HAI, where Li Feifei works, has released 8 versions of the AI Index report since 2017, aiming to track activities and progress in the field of artificial intelligence and promote discussions on data-driven artificial intelligence. Committed to providing accurate, rigorous, and global AI data and insights for policy makers, researchers, business executives, and the public.

In February 2020, Li Feifei was elected as a member of the National Academy of Engineering in the United States for his outstanding contributions in establishing large-scale machine learning and visual understanding knowledge bases. In October of the same year, she was elected as a member of the National Academy of Medicine in the United States. In April 2021, she was elected as a member of the American Academy of Arts and Sciences.