Hello BeagleBoard,
This is Nishant Malpani, a fourth-year integrated Master’s student, majoring in Electronics and Communications (ECE) engineering, at International Institute of Information Technology, Bangalore, India (IIIT-B).
Having worked on various Deep Learning projects, I have come to realise the lack of computing power when it comes to deploying such a model on a dedicated development board like Raspberry Pi 2/3 or STM32F4-Discovery. This challenge-driven attraction towards Deep Learning techniques with my curiosity for embedded systems’ software always intrigued me into working in the field of ‘Systems for ML’. Having said that, BeagleBoard’s project on ''YOLO models on the X15/AI" for GSoC 2020 has caught my interest and I’d love to contribute to such an interdisciplinary project. Excuse me for my selfish reasons for choosing BeagleBoard, but I believe that I can acquire a lot of learning opportunities with this organization.
Formally, I’ve done courses on 'Linux Device Driver Development’, ‘Operating Systems’, ‘ARM Architecture’, ‘Virtual Machines’ and currently enrolled in a course on ‘Real-Time Operating Systems’ at my institute. On the Machine Learning front, I’ve completed courses like ‘Mathematics for Machine Learning’, ‘Computer Vision’ and ‘Deep Reinforcement Learning’. My journey with the open-source community has just begun; I’ve started to delve myself to provide drivers-related patches on the Linux kernel’s staging branch. One may call me a ‘newbie’ when it comes to open source contributions but I believe I have it in me to take the initiative to contribute in open source projects for long-term. With this being said, I’d be delighted to start contributing to BeagleBoard and hope to add a great value to the team.
I’ve completed the introductory task on cross-compilation and the puIl-request can be found here. I believe I have a sufficient idea of how YOLO, and object detectors in general, work. I have cloned the beaglebone-ai repository to try to understand the codebase.
I’d appreciate if someone could direct me to a good starting point/task for the aforementioned project.
Personal details:
- GitHub: layman-n-ish
- LinkedIn: Nishant Malpani
- Email: nish.malpani25@gmail.com
I’d be more than happy to provide you all with any other details. Thank you for your time.
With regards,
Nishant Malpani
Hello BeagleBoard,
This is Nishant Malpani, a fourth-year integrated Master’s student, majoring in Electronics and Communications (ECE) engineering, at International Institute of Information Technology, Bangalore, India (IIIT-B).
Having worked on various Deep Learning projects, I have come to realise the lack of computing power when it comes to deploying such a model on a dedicated development board like Raspberry Pi 2/3 or STM32F4-Discovery. This challenge-driven attraction towards Deep Learning techniques with my curiosity for embedded systems’ software always intrigued me into working in the field of ‘Systems for ML’. Having said that, BeagleBoard’s project on ''YOLO models on the X15/AI" for GSoC 2020 has caught my interest and I’d love to contribute to such an interdisciplinary project. Excuse me for my selfish reasons for choosing BeagleBoard, but I believe that I can acquire a lot of learning opportunities with this organization.
Formally, I’ve done courses on 'Linux Device Driver Development’, ‘Operating Systems’, ‘ARM Architecture’, ‘Virtual Machines’ and currently enrolled in a course on ‘Real-Time Operating Systems’ at my institute. On the Machine Learning front, I’ve completed courses like ‘Mathematics for Machine Learning’, ‘Computer Vision’ and ‘Deep Reinforcement Learning’. My journey with the open-source community has just begun; I’ve started to delve myself to provide drivers-related patches on the Linux kernel’s staging branch. One may call me a ‘newbie’ when it comes to open source contributions but I believe I have it in me to take the initiative to contribute in open source projects for long-term. With this being said, I’d be delighted to start contributing to BeagleBoard and hope to add a great value to the team.
I’ve completed the introductory task on cross-compilation and the puIl-request can be found here. I believe I have a sufficient idea of how YOLO, and object detectors in general, work. I have cloned the beaglebone-ai repository to try to understand the codebase.
I’d appreciate if someone could direct me to a good starting point/task for the aforementioned project.
You might have seen the beaglebone-ai repo is really just the hardware. As mentioned on another thread, check the IRC chat log http://logs.nslu2-linux.org/livelogs/beagle-gsoc/ for some relevant discussion. The codebase is the TIDL stuff on git.ti.com.