Nvidia Software Engineer Intern Perception - Autonomous Vehicles 

China, Shanghai 

226900887

DL models, including data mining, AV data pre-processing and post-processing pipelines. Our work will be fun and veryexciting, and

be doing:

• Will mainly focus on building automatic pipelines of GT generation for AV DL models of perception, plan and control, including data mining, data processing, auto-labeling, ground truth generation and diagnosis, and performance optimization of the pipelines running on cloud service. We need to have passion for both deep learning and autonomous driving domain!

• Work closely with AV perception core team, AV production team, AI infrastructure and MLOps engineers to accelerate and optimize the AV perception DNN development and deployment, and achieve the high quality and reliability of AV perception in NVIDIA Drive products.

What we want to see:

• Pursuing MS or PHD candidate in Computer Science, Computer Engineering, Electrical Engineering, Robotics, or related majors

• Excellent programming, software design, and testing skills in Python, C++, CUDA, or similar languages.

• Good knowledge in either of the following areas:

• 3D Computer Vision, Self-Driving-Cars, Robotics, SLAM, Deep Learning, GPU technology.

• Good understanding of system architecture,CPU/GPU/Memory/Storage,everything related to performance optimization.

• Solid knowledge of Linux programming and should be familiar with working under Linux environments.

• Fluent in both oral and written English.

Ways to stand out from the crowd:

• Background in computer vision, machine learning, deep learning, statistics

• Knowledge of LLM optimization, acceleration and application. Experience with NVIDIA Triton will be a big bonus

• Knowledge of object detection/tracking, multi-sensor fusion such as lidar or camera, and timesynchronization.

• Experience with large data exploration, analysis and visualization with Python and SQL.

• Experience with deep learning models training with popular frameworks like Tensorflow, Pytorch, Caffe and model deployment as well as k