About WM Motor:
Weltmeister Motor is one of the leading electric vehicle companies. With offices in Shanghai, Chengdu, Wenzhou and the bay area, we are building a high-performance team that is excited by complex engineering challenges and is passionate about making transportation safer, more affordable and accessible for all.
The team you will be working with:
The machine learning team is responsible for designing and implementing algorithms to perceive and predict the environment from sensor data. The technical lead will work with a cross-functional team as well as outside suppliers and bring the system into production.
- Work with the teams to define the technical requirements
- Lead a team of engineers to research, design, implement, test and deploy robust and high-performance machine learning systems for autonomous driving
- Design and implement large-scale machine learning pipelines
- Design and develop models to detect lane line, pedestrian, vehicle, traffic signal, and free space, etc.
- Design and develop models for predicting the behavior of the other actors
- Research and explore new technology to improve the systems
- Knowledge of the theory and practice of deep learning and its recent advancements
- Knowledge of modern GPU architectures
- Proficient in creating labelling documents to communicate the requirements with the third-party labelling company. Integrate the pipeline with the cloud infrastructure
- Proficient in training and evaluating deep neural networks to detect lane lines, vehicles, traffic signals and free space in practice
- Proficient in optimizing the tradeoff between speed and performance of the networks
- Proficient in isolating, documenting, and tracking issues systematically
We appreciate if you have:
- Experience of developing machine learning products
- Experience of C/C++11
- Experience of managing GPU clusters for training deep networks
- Experience of compressing and porting the neural networks to embedded platforms
- Knowledge of fundamentals: discrete/continuous optimization methods, supervised/unsupervised learning, generative/discriminative models.
- Knowledge of classical computer vision techniques and camera models
- Knowledge of CUDA or OpenCL programming