- The Computer Vision and Pattern Recognition conference has kicked off in Salt Lake City, Utah!
- NVIDIA has laid down the first marker with four major updates – TensorRT 4, Apex, NVIDIA DALI and Kubernetes on NVIDIA’s GPUs
- Codes for most of these are open sourced and available on GitHub
NVIDIA has emerged as one of the leading organizations in the machine learning and deep learning space. We have previously seen some breakthrough software from them in this field – from a robot that can copy and execute human actions to an open source Python library that makes anyone an artist.
And now they have announced a slew of machine learning tools at the Computer Vision and Pattern Recognition Conference (CVPR) in Utah. CVPR is an annual machine learning conference which sees the top minds in the ML and DL industry come together to discuss and present the latest tools and research to the community.
These latest tools by NVIDIA include TensorRT 4, Apex, NVIDIA DALI (data loading library) and Kubernetes on NVIDIA’s GPUs. We have summarized each of these four tools in this article for you!
- New Recurrent Neural Network (RNN) layers for Neural Machine Translation apps
- New Multilayer perceptron (MLP) operations and optimizations for Recommender Systems
- Native ONNX parser to import models from popular deep learning frameworks
- Integration with TensorFlow
Apex is an open source PyTorch extension that helps data scientists and AI developers maximize the performance of their deep learning training process on NVIDIA’s own Volta GPUs. It has been inspired by state-of-the-art techniques like sentiment analysis, translational networks, and image classification.
You need to have CUDA 9, PyTorch 0.4 and Python 3 installed before using Apex. It’s still in alpha mode, meaning there is a lot more to come before the first official full launch. Since it’s an open source release, the code is available for everyone to download on GitHub.
Kubernetes on NVIDIA GPUs
NVIDIA DALI and NVIDIA nvJPEG
DALI, short for Data Loading Library, aims to tackle the image recognition and computer vision field. It leverages the power of its own GPUs to decode images at a much greater speed than ever before.
nvJPEG on the other hand aims to support “decoding of single and batched images, color space conversion, multiple phase decoding, and hybrid decoding”. These two are tied together as DALI relies heavily on nvJPEG for that acceleration in speed.
DALI is an open source release and you can download the code on GitHub.
You can read about each of these tools in more detail on NVIDIA’s Developer Blog using the below links:
Our take on this
First, I highly recommend everyone check out the CVPR website. As a data scientist (or an aspiring one), this conference is a gold mine for you. There are tons of updates and research papers being presented at this conference by the top minds in the machine learning field. Go through the website, check out the resources and you are sure to learn a lot of new things you did not know about before.
Coming to NVIDIA, I can see Apex becoming a hit with deep learning users. Ever since PyTorch hit the open source market, it has seen a rapid adoption rate and combined with Apex, it could be a potential game changer.
Subscribe to AVBytes here to get regular data science, machine learning and AI updates in your inbox!
You can also read this article on our Mobile APP