Introduction to Convolutional Neural Networks using TensorFlow and Keras
summary
Uses TensorFlow and Keras to classify German speed‑limit signs from a tiny dataset (~379 images, six labels) by resizing everything to 64x64 images, normalizing and trying architectures: simple dense nets fail, convolutional nets reach about 90% test accuracy (training ~96%) but real‑world photos hit only 5/15 correct. Training runs on Microsoft/Azure notebooks, uses RMSprop, dropout, max‑pooling, batch‑norm; experiments include Inception‑style towers and residual (very small) models; model files are saved and shared (transfer.sh) for live testing.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 10:48Setting up the environment and dependencies
- 21:49Neural network basics
- 33:59Model compilation
- 42:30Testing and improving
- 53:32Convolutional neural networks
- 1:06:27Training and evaluation
- 1:19:18Advanced architectures
- 1:25:26Model challenges
- 1:30:41Q&A
related talks
51mA Practical Introduction to Convnets for Image Classification
57mTensorflow, deep learning and modern convolutional neural nets, without a PhD
01h 29mHands-on introduction to Deep Learning with Keras and Tensorflow
01h 06mDeep Learning with TensorFlow
37mCode First, Math Later: Learning Neural Nets Through Implementation and Examples
55mA Tour of Deep Learning With C++
32mA word is worth a thousand pictures: Convolutional methods for text
01h 29mComputer Vision: An (Un?)Expected Journey, with Keras and Tensorflow