- Investigate frameworks for neural architecture search (NAS), supported by novel profiling methods, extending them with mixed precision capabilities, and investigate search space and optimization technologies
- Extending the frameworks with multi-objective optimization for qualities such as latency, memory usage, energy-efficiency, applying this to the families of resource-constrained inference devices
- Extending these frameworks for on-device and federated learning
- Extending hardware-aware NAS for next generation -targets up to the point of intelligently designing and configuring novel configurable deep learning accelerators
- The methods are applied to vision, radar, audio and other time-series-data for above-mentioned industries
- Integration of the investigated methods and neural networks in larger applications, open source frameworks and compelling demonstrators
Department & Project:
You would be based in our client's Automotive System Innovations (ASI) department which is part of the Chief Technology Office.
They drive innovation on system level for the Automotive businesses in applications like highly automated and safe driving, audio, radar systems, in-vehicle networking, artificial intelligence, battery management systems, as well as mobile robotics.
The work is performed in a team of AI-experts, embedded in world-wide AI-activities. The team’s current scope in research and advanced development is to research and evaluate methods for efficient machine learning (ML) inference while fully exploiting the HW/SW characteristics of all AI platforms, targeting multi-objective optimization.
They support the BLs with deep know-how to embed these ML-methodologies in HW/SW AI products. They stay on top of and utilize today’s state of the art developments in the academic and open source communities, and participate there-in through partnerships, PhD- and master-student programs, publications etc.
We seek to expand the team with ambitious embedded AI architects/engineers.
- University degree: MSc, PhD or PDEng in a technical specialism, like Computer Science or equally relevant.
- 3+ years of experience in (software) engineering, with significant exposure/involvement with Machine Learning / AI required
- Affinity and experience with embedded processors, software and NN accelerators required.
- Broad experience with embedded software architectures, build systems, version control systems required.
- Broad experience with Operating systems GNU/Linux, embedded systems, development boards, and processors, and SW competencies as listed below required.
- Excellent communication skills in English (verbal /written) required. Experience in working in/with multi-site and multi-cultural projects/teams preferred.
- Knowledge of electronics design which helps to solve issues on the intermediate HW-SW layer preferred.
- Flexibility in working with AI frameworks (TensorFlow, PyTorch), preferably via Python and C++ interfaces required.
- Understanding of AI toolchains, deployment, portability and inference engines (CUDA, TensorRT, TFLite, ONNX, etc.) preferred.
- Familiarity with setting up and maintaining related development environments (docker, TensorBoard, performance profiling, ClearML, etc.) required.
- Knowledge of build systems (YOCTO, OpenEmbedded, etc.) beneficial, working with cross-compilation toolchains for ARM preferred.
- Solid programming experience of C, C++, Python and Bash programming languages on Linux systems required.
- Experience with on-device learning and federated learning appreciated.
- Familiarity with future technologies for extreme efficiency such as in-memory, analog, and neuromorphic computing beneficial.
Please contact: firstname.lastname@example.org