We reproduce the latest Fast.ai/DIUx's ImageNet result with a single 8 Turing GPUs (Titan RTX) server. It takes 2.36 hours to achieve 93% Top-5 accuracy.
This blog tests how fast does ResNet9 (the fastest way to train a SOTA image classifier on Cifar10) run on Nvidia's Turing GPUs, including 2080 Ti and Titan RTX. We also include 1080 Ti as the baseline for comparison.
This blog gives a brief introduction on the history of object detection, explains the idea behind Single-Shot Detection (SSD), and discusses a number of implementation details that will make-or-break the performance. Tensorflow implementation is also provided.
This tutorial will explain the main ideas behind some of the popular networks for image segmentation, including fully convolutional networks and U-Net. Tensorflow implementations of these networks are provided.
This tutorial covers the basic of hyper-parameter tuning in the context of image classification. We use ResNet32 and CIFAR10, and showcase how to find the optimal optimizer and learning rate using the tuner function in Tensorflow.
This tutorial demonstrates how to use a pre-trained model for transfer learning. The networks used in this tutorial include ResNet50, InceptionV4 and NasNet. The dataset is Stanford Dogs. Tensorflow implementation is provided.
This tutorial will walk you through the steps of building an image classification application with TensorFlow. We will also introduce you to a few building blocks for creating your own deep learning demos.
...