Deep learning Workshop for Satellite Imagery - Training & Prediction (Part 2/3)
If your interested into deep learning for the satellite images, this full hands-on coding workshop is best resources for you. The full workshop is divided into following 3 parts:
1. Data Processing for the Satellite Imagery
Data Preparation
Satellite Data Processing
2. Deep Learning with Satellite Image Data
Deep Learning Model Training
Local Debugging
Prediction with model performance
3. Advance Deep Learning with Satellite Image Data
Remote debugging with Weights & Biases
Model Saving and Reloading with custom metrics and loss function
Activation/Gradients outputs with heatmap,
Model Deployment application and Model Serving app on Hugging Face
GitHub Resources:
https://github.com/prodramp/DeepWorks...
▬▬▬▬▬▬ ⏰ TUTORIAL TIME STAMPS ⏰ ▬▬▬▬▬▬
(00:00) Content Starts
(00:30) What is covered?
(04:50) Part 1 Refresher
(06:30) Part 2 Coding Starts
(09:45) Jaccard Index
(15:15) U-Net Model Architecture
(20:01) DL Network Coding Starts
(36:45) Model Metrics
(38:20) Code Debugging 1
(42:45) Custom Loss function
(49:10) Model Training
(50:55) Code Debugging 2
(55:55) Model Metrics/Eval
(1:02:45) Model Prediction
(1:02:45) Model Prediction
(1:10:50) Save/Export Model
(1:12:30) Model Network Plot
(1:14:20) Diagnostics with Callback
(1:27:00) Model Visualization with Netron
(1:32:00) Source Code at GitHub
Connect
------------------
Prodramp By Avkash (@prodramp)
Website - https://prodramp.com
LinkedIn - / prodramp
GitHub- https://github.com/prodramp/
AngelList - https://angel.co/company/prodramp
Facebook - / prodramp
Content Creator: Avkash Chauhan (@avkashchauhan)
/ avkashchauhan
/ avkashchauhan
Tags:
#tensorflow #satelliteimagery #deeplearning #keras #prodramp #avkashchauhan
