Jiaming Li Email & Phone Number

Palo Alto, California, United States

Responsible for utilizing the cloud machine learning platform to improve Varian’s sales strategies and developing/maintaining Varian web applications.

Marietta, Georgia, United States

•Invented Temporal Spatial Landmarks Feature Map (TSLFM) that encoded the temporal and spatial facial landmarks changes in the 2D matrix for facial expression recognition in continuous video frames. The experiment indicated that the proposed encoding method achieved 91.27% accuracy on classifying 7 different facial expressions in CK+ dataset using the CNN model.•Put forward Stratified Experience Replay that improved the training efficiency of Deep Q-learning Networks (DQN) by 20.70% learning on CartPole-v1 environment of OpenAI Gym.•Proposed self-adaptive Seep Up Robust Features (SURF) that achieve real-time image registration in video sequence.•Designed a Convolutional Neural Network based on pre-trained VGG-16/19 model for Coding Tree Unite (CTU) partition pattern predication that improves the encoding efficiency of HEVC/H.265 video compression standard.•Presented a three-level classifier structure using CNN model to predict the depth of various size of Coding Unite (CU) that saves up to 73.80 % encoding time of HEVC with negligible 2.645% average Bjøntegaard delta bit-rate increase.

Downers Grove, Illinois, United States

Responsible for developing and maintaining a one-stop-shop cloud and DevOps resource management system using contemporary AI technologies and integrating social login functionality to Syngenta web application in an agile team. •Automated the Glue tables creation process with Python script that generates table schemas according to the Postgres table structure stores in AWS RDS and creates the tables using AWS CLI commands. •Deployed the Web Crawler that collects the AWS and DevOps tools (OpenShift, Service Now, Jira, etc.) resource usage and status information by calling the REST APIs. •Utilized Apache Spark to parallelize the data extract, transform and load (ETL) process. •Visualized the data to interactive dashboards using AWS QuickSight and delivered them to the web application. •Built a Chatbot application that provides the cloud resource information using AWS Lex to recognize the natural language and AWS Lambda to fulfillment the business logic. •Integrated social login to SpringBoot web application using Okta as a third-party identity provider. And deployed the project to the OpenShift container.

Master'S Degree, Computer Science

Bachelor'S Degree, Computer Science

Bachelor'S Degree, Computer Science