You're invited to join me as I speak on Image Captioning in Pytorch
at NVIDIA’s GPU Technology Conference March 17-21 in Silicon Valley.
In this lab, we'll walk through an example of image captioning in Pytorch. We'll begin by downloading and training on the coco image dataset, review data augmentation with cropping, rotating, flipping and resizing images. We'll then build a vocabulary for the image annotations and encode the sequences as captions. Our model for this task includes a pretrained encoder CNN to compute features on each image and the decoder RNN which handles the captioning. I'll show you the results of training and evaluation from this model including automatically captioning images not included in the training set.
Use discount code GMSGTC good for 20% off any pass.
Are you part of the LatinX in AI or Black in AI Community? Apply for the "I AM AI" Scholarship Grant, sponsored by NVIDIA!
Hope to see you at GTC 2019!