Object detection models built with TensorFlow. There are multiple variations of models, some built for efficiency (mobile inference) and some for accuracy
An automatic speech recognition model built for TensorFlow.
A PyTorch implementation of the Google Tacotron architecture for natural sounding text to speech synthesis, trained using the LJ speech corpus.
Facebook AI Research (FAIR) implementation of object recognition for Caffe2 trained on the COCO dataset.
Various models trained for a range of tasks