都柏林大学2025 - PhD Research Scholarship in Dynamic AI Model Compression

Applications are invited for a fully funded 4-year PhD position at UCD, Ireland.
Host Institution: University College Dublin,  Ireland  (QS Ranking) (THE Ranking) (US News)
Research Advisor: Dr. Deepu John
Scholarship: Full Tuition Fee waiver + € 25,000 per year stipend + Variable TA Pay
 
Background
This project will investigate advanced techniques for compressing machine learning models and developing dynamic neural networks for edge AI applications. The aim is to optimize the performance and efficiency of complex models, enabling their deployment in resource-constrained environments such as IoT devices.
Traditional machine learning models often require substantial computational resources, power, etc, posing challenges for deployment in edge devices. This project will address these challenges by developing methodologies for dynamic model adaptation, ensuring efficient performance with minimal resource consumption. The hired candidate will be expected to build upon our prior works in dynamic AI model compression.
1.       https://arxiv.org/abs/2403.01695
2.       https://arxiv.org/abs/2403.17726
Key Objectives

• Develop novel model compression techniques      capable of dynamic adaptation.
• Design dynamic neural networks that can adapt      their structure and parameters in real time based on available      computational resources.
• Validate these techniques in real-world embedded,      edge AI systems, such as IoT devices
• Evaluate these techniques on RISC-V SoCs developed by our team members

 
The project will involve collaboration with multi-disciplinary researchers working on RISC-V SoCs, AI accelerator design, IoT/Edge Sensors, etc. The position also provides the opportunity to gain valuable experience in design/testing of edge systems. The successful candidate will get opportunities to present their work at international conferences and workshops and publish the research in high-impact journals.
 
Applicants for this position should ideally have:
·Bachelor's/Master’s degree in Electrical /Computer Engineering or equivalent.
·Experience in machine learning and model compression techniques.
·Strong analytical and programming skills in Python/Matlab and/or embedded C/C++
·Proficiency in ML frameworks (e.g., TensorFlow, PyTorch or other similar)
·Experience edge AI, dynamic neural networks and/or embedded systems.
·A strong publication record in leading conferences and journals is desirable.
·Good interpersonal and English language skills (IELTS/ToEFL or willing to take)
·Enthusiasm and willingness to explore new ideas and concepts.
Application open: Immediate, Position will be closed as soon as a suitable candidate is identified.


How to Apply
The project will be supervised by Dr Deepu John. Applications containing a cover letter, curriculum vitae, and degree transcripts should be sent to deepu.john@ucd.ie with the subject line PhD Application 2025. The position will remain open until a suitable candidate is identified. Applications will be reviewed on a rolling basis. Only shortlisted candidates will be contacted.

upupup，在线顶贴