# 296: Train Me Later

- Show: [Embedded.fm](https://hardwarepodcasts.com/podcast/embedded-fm)
- Published: 2019-07-25
- Length: 45m
- Episode: 296
- Page: https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later
- Listen: https://makingembeddedsystems.libsyn.com/296-train-me-later

## Summary

Shruthi Jaganathan, a high school intern at NVIDIA, describes building the Green Machine, a computer vision system using a Jetson TX2 and later a Jetson Nano to sort cafeteria waste. She explains how her team trained a neural network on 5,000 labeled images of trash, using TensorFlow and SSD MobileNet to classify items as recyclable, compostable, or trash. The system uses a camera and projector to display color-coded boxes on items, and Jaganathan details the challenges of calibration, data augmentation, and porting the software to the smaller Jetson Nano using JetCam.

## Chapters

- [0:06](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=6) Introduction: Shruthi Jaganathan and NVIDIA Internship
- [5:00](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=300) The Green Machine: Jetson TX2 and Computer Vision
- [7:21](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=441) Data Collection and Labeling Challenges
- [12:03](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=723) TensorFlow, SSD MobileNet, and Multi-class Detection
- [14:51](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=891) Camera Calibration and QR Code Alignment
- [20:03](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=1203) Porting to Jetson Nano and JetCam
- [26:53](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=1613) Distinguishing Compostable vs. Reusable Utensils
- [28:42](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=1722) Data Augmentation and Handling Projector Light
- [35:27](https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later#t=2127) Open Source Code and Future Plans

## Topics

- [Embedded firmware](https://hardwarepodcasts.com/topic/embedded-firmware)
- [AI & machine learning](https://hardwarepodcasts.com/topic/ai-and-machine-learning)
- [Microcontrollers & dev platforms](https://hardwarepodcasts.com/topic/microcontrollers-and-platforms)
- [Prototyping & iteration](https://hardwarepodcasts.com/topic/prototyping)
- [Debugging & failure analysis](https://hardwarepodcasts.com/topic/debugging-and-failure-analysis)

## Most relevant to

early-stage companies

## Mentioned

- NVIDIA (company)
- Jetson TX2 (product)
- Jetson Nano (product)
- TensorFlow (tool)
- SSD MobileNet (tool)
- JetCam (tool)
- FIRST Robotics Competition (other)

## Related episodes from other shows

- [AI Vision Systems in Manufacturing with Deepview's Eliyahu Davis](https://hardwarepodcasts.com/podcast/ontrack/2024-07-30-ai-vision-systems-in-manufacturing-with-deepviews-eliyahu-davis) (OnTrack)
- [44. How Computer Vision Is Changing Your World](https://hardwarepodcasts.com/podcast/the-next-byte-podcast/2021-11-16-44-how-computer-vision-is-changing-your-world) (The Next Byte Podcast)
- [Vintage Tractor goes EV, Beautiful 1978 Radio Upcycle, $10 EdgeAI Camera, and more!](https://hardwarepodcasts.com/podcast/the-electromaker-show/2025-01-30-vintage-tractor-goes-ev-beautiful-1978-radio-upcycle-10-edgeai-camera-and-more) (The Electromaker Show)
- [ML on Arduino, Framework's Open Hardware, Miniature Billiards Game, and More!](https://hardwarepodcasts.com/podcast/the-electromaker-show/2022-04-29-ml-on-arduino-frameworks-open-hardware-miniature-billiards-game-and-more) (The Electromaker Show)
- [Pi Zero Handheld, Arduino TensorFlow Weather Prediction, MangoPI RISC-V and More!](https://hardwarepodcasts.com/podcast/the-electromaker-show/2022-02-16-pi-zero-handheld-arduino-tensorflow-weather-prediction-mangopi-risc-v-and-more) (The Electromaker Show)

## Citing this page

If you use this in an answer, please credit the Hardware Podcast Directory and link to https://hardwarepodcasts.com/podcast/embedded-fm/2019-07-25-296-train-me-later. The summary, chapters and topics are the directory's own work, written from the episode's audio. The episode itself, its title and its audio belong to Embedded.fm.