Long Chen
Software Engineer at LiveKit
Hunan, China
Summary
🤩
Rockstar🎓
Top SchoolLong Chen is a software engineer with 11 years of experience specializing in computer vision and real-time multimodal systems, currently building LiveKit Agents' backend integrations. He has led people-tracking and multi-view 3D pose projects as a tech lead at AiFi and shipped research-driven solutions during internships at Apple and AiFi. His open-source work includes CUDA-optimized ROI pooling for a Faster R-CNN PyTorch implementation and YOLOv2 adaptations, showing deep hands-on expertise in accelerating detection pipelines. Comfortable across research and production, he pairs a Tsinghua PhD-level background with practical product deliveries like shelf detection and human activity recognition. A subtle strength is his ability to bridge low-level performance engineering (CUDA layers) with high-level realtime API and speech integrations for multimodal agents.
11 years of coding experience
6 years of employment as a software developer
Bachelor's degree, Electronic Information, Bachelor's degree, Electronic Information at Huazhong University of Science and TechnologyDoctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Tsinghua University
Github Skills (17)
pytorch10
openai-api10
python10
machine-learning10
mask-rcnn10
darknet10
cuda10
object-detection10
computer-vision10
faster-rcnn10
agent8
ai8
kubernetes-pods5
go5
dockers5
Programming languages (9)
Github contributions (5)
longcw/yolo2-pytorch
Feb 2017 - Sep 2021
YOLOv2 in PyTorch
Role in this project:
ML Engineer
ML EngineerContributions:46 commits, 8 PRs, 35 pushes in 4 years 7 months
Contributions summary:Long primarily contributed to implementing and integrating a YOLOv2 object detection model in PyTorch. Their commits include loading pre-trained weights, modifying the model's architecture, and integrating a demo for object detection. They focused on adapting the model for a specific dataset and configuring the preprocessing and post-processing steps. The user also demonstrated skills in data loading and model evaluation.
deep-learningpytorchobject-detectionyolov2darknet
longcw/faster_rcnn_pytorch
Jan 2017 - Sep 2021
Faster RCNN with PyTorch
Role in this project:ML Engineer
ML EngineerContributions:26 commits, 3 PRs, 32 pushes in 4 years 8 months
Contributions summary:Long primarily contributed to the development of a Faster R-CNN implementation using PyTorch. Their work included modifications to core modules such as `faster_rcnn.py`, `roi_pool.py`, and `roi_pooling_cuda.c`, indicating a focus on the model's architecture and performance. A significant amount of the commits concentrated on implementing and debugging a CUDA-based ROI pooling layer, crucial for accelerating the object detection process. The user also addressed a bug related to handling scenarios with no valid detections.
pytorchfaster-rcnnfasterrcnncomputer-vision
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.