Responsibilities

• Responsible for the R&D and mass production implementation of in-vehicle ADAS and DMS algorithms

• Undertake end-to-end algorithm engineering, including data collection and cleaning, model training and tuning, distillation/ quantization/ pruning for lightweighting, embedded on-device deployment, inference performance optimization, in-vehicle integration testing and mass production iterations

• Collaborate with engineering teams to complete algorithm SDK packaging, cross-chip platform adaptation, and optimization of power consumption and real-time performance

• Conduct preliminary research into in-vehicle AI technologies, exploring the lightweight implementation and integrated application of large language models and multimodal large models in in-vehicle perception and behaviour understanding, and carry out technical solution validation and iteration

• Be responsible for coordinating and advancing resources of the Companyt’s Research Center; possess a broad perspective and demonstrate business acumen (marketing/product/project management)

• Collaborate with universities on research and project partnership

Requirements

• Master’s degree in Computer Science, Artificial Intelligence, Automation, Electronic Information, Machine Vision or related fields

• At least 5 years of experience in the mass production R&D of in-vehicle vision algorithms

• Solid foundation in traditional computer vision; proficiency in core algorithms such as object detection, instance segmentation, key point detection and classification

• Experience in the mass production implementation of ADAS, DMS, 360-degree surround view or SLAM; with proficiency in algorithmic optimisation for complex driving conditions such as strong glare, backlighting, night-time, occlusion and motion blur

• Proficiency in the engineering deployment of inference frameworks such as NPU chips, MNN and NCNN; and in lightweighting techniques including knowledge distillation, INT8/INT4 quantisation and structural pruning

• Proficient in Python and the PyTorch deep learning framework; possesses solid C++ engineering development skills, as well as SDK encapsulation and cross-platform adaptation capabilities

• Possess strong learning and technical transfer capabilities; proactively track developments in large-scale model technology; able to quickly master LoRA/QLoRA fine-tuning, multimodal model lightweighting and on-device deployment of large-scale models; possess composite capabilities in the integration and implementation of traditional computer vision and large-scale models

• Excellent  communication , presentation and interpersonal skills

• Fluent English, Cantonese and Mandarin

• Frequent travel to PRC is necessary

Interested parties, please send your resume with expected salary to [email protected]. You can also click on the website www.easyjobhk.com for more details about the company