Workshop material for "Deep learning for computer vision and activity recognition lab" held during the 2018 JHU Summer School on Human Language Technology .
Lab notebooks:
- Tensorflow MNIST Tutorial 2.ipynb : Introduction to image classification using Tensorflow. Based on Tensorflow's official tutorial Deep MNIST for Experts and the jupyter notebook TensorFlow Tutorial.
- video_classification_i3d_tf.ipynb: Introduction to video classification using Tensorflow. This notebook builds upon the demo script provided by the authors of the the paper "Quo Vadis,Action Recognition? A New Model and the Kinetics Dataset".
- temporal_action_segmentation_tf.ipynb: Train a recurrent neural network on the Breakfast dataset for the task of temporal action segmentation and recognition.
Notebooks include instructions for executing them on Google Colaboratory.