Image Recognition DL WORKSHOP
summary
An end-to-end workshop showing how to classify individual buildings from aerial and satellite imagery using a Google Colab notebook. It covers dataset layout with a held-out validation set, real-time image augmentation and generator pipelines, training via transfer learning from ImageNet-pretrained CNNs (VGG/Inception/ResNet), checkpoint callbacks and dashboarded metrics, and a submission workflow for the PyData challenge (examples include damage mapping for Sint Maarten and material classification).
AI-written summary · about
chapters (13)
- 0:00Workshop intro
- 4:52Challenge setup
- 8:00Colab notebook
- 11:59Deep learning pipeline
- 14:36CNN for image classification
- 17:49Fine-tuning models
- 21:18Real-time data augmentation
- 23:29Model architectures
- 25:48Custom classification layer
- 28:10Training callbacks
- 31:45Dashboard and metrics
- 34:14Class imbalance
- 36:28Submitting predictions
related talks
01h 39mGoogle Colaboratory and Python: Deep Learning basics for CV tasks
01h 36mFrom Maps to Models: An End-to-End Journey in Geospatial Machine Learning with Python
01h 29mComputer Vision: An (Un?)Expected Journey, with Keras and Tensorflow
01h 30mFundamentals of Deep Learning Workshop
49mKeynote: From Zero to ML on Google Cloud Platform
01h 27mHands-On Computer Vision with PyTorch
02h 41mHello TensorFlow: 3 Workshops to Get Started with TensorFlow 2.0
35mDeconstructing Deep Learning