Mingju He

I am a distinguished Computer Engineering PhD from Stevens Institute of Technology with a focus on machine and deep learning, notably in signal/modulation classification. With hands-on experience at Xsense.ai predicting vehicle movements using the Waymo Open Dataset, I combine academic excellence with practical expertise. I also have published multiple papers and recognized as an outstanding teaching assistant, I seek roles to further innovate in applied science and machine learning.

Occupancy and Flow Prediction (September 2022 - December 2022)

• Position: Deep Learning Research Intern at Xsense.ai, San Diego, CA
• Tasks & Achievements:
  • Used the Waymo Open Dataset (103,354 segments) to predict future vehicle movement and occupancy.
  • Extracted features from TF.record files, reducing data size for efficiency.
  • Developed a Tensorflow data loader for efficient input.
  • Designed a Transformer-based model with ResNet and custom CNN layers for sequential data.
  • Evaluated model, achieving improvements in AUC, IOU, and EPE metrics by 5.6%, 6.4%, and 2.5% respectively over a CNN model.
  • Cut training time by 50% using early stopping and efficient data loading.
  • Visualized driving scenarios and emphasized the model's potential to reduce risks and accidents.

Stevens Institute of Technology Hoboken, NJ

Ph.D in Computer Engineering;

GPA: 4.00/4.00

August 2018 – May 2023