Experience

Axelera AI

• Netherlands
• Semiconductor Manufacturing
• 1 - 100 Employee

• Deep Learning Research Engineer

  • Apr 2022 - Present

  From April 2023, I extend to the topic of Neural architecture search (NAS). Develop training-free NAS algorithm with latency and FLOPs constrains. Let Axelera's AI chips say 'yes' to the searched tiny model. From April 2022, research on network pruning: measure the power efficiency for different pruning granularities, develop new algorithm, and output one accepted ICCV2023 paper. Main research interests are network pruning, neural architecture search and network quantization. From April 2023, I extend to the topic of Neural architecture search (NAS). Develop training-free NAS algorithm with latency and FLOPs constrains. Let Axelera's AI chips say 'yes' to the searched tiny model. From April 2022, research on network pruning: measure the power efficiency for different pruning granularities, develop new algorithm, and output one accepted ICCV2023 paper. Main research interests are network pruning, neural architecture search and network quantization.



Delft University of Technology

• Netherlands
• Research Services
• 700 & Above Employee

• Guest Researcher

  • Feb 2022 - Present

  Optimization of binarized neural network: output one accepted ICLR2023 publication with my supervised master student.

• Researcher PHD Student

  • Feb 2018 - Jan 2022

  Main research interest is focusing on deep compression, including binary neural network, network pruning, and deep hashing for semantic retrieval!