Deep Learning Weekly Issue #132

Graphcore coming to Azure, Nvidia's DeepStream, TF 2.0 hackathon, BodyPix for tf.js, and more...

Hey folks,

This week in deep learning we bring you news of Graphcore accelerators in Azure, a TensorFlow 2.0 hackathon, Nvidia’s DeepStream SDK for edge IoT devices, and a neural network for foveated reconstruction in VR / AR from Facebook.

We also bring you a new body part estimation model from the TensorFlow team, a write-up of Kuzushiji recognition for ancient Japanese documents, a small course on exploiting and hacking neural networks, the official release of MobileNetV3, a new PyTorch library to accelerate 3D deep learning research from Nvidia, and recent work on few shot learning.

As always, happy reading and hacking. If you have something you think should be in next week's issue, find us on Twitter: @dl_weekly.

Until next week!


Advanced machine learning helps Play Store users discover personalised apps

DeepMind explains how their LSTM model improves PlayStore rankings.

NVIDIA DeepStream SDK for IoT and Real-Time Streaming Analytics Debuts on Microsoft Azure Marketplace

DeepStream, an SDK containing hardware-optimized DL frameworks and pre-trained models is now available on the Azure Edge IoT marketplace.

Graphcore’s AI accelerator chips launch on Microsoft Azure

The first generation of neural network accelerators is scaling up.

TensorFlow 2.0 hackathon with $150,000 in prizes.

TensorFlow 2.0 hackathon submissions are due by Dec. 31st.


How Machine Learning Can Help Unlock the World of Ancient Japan

A neat write-up on training models to perform Kuzushiji recognition for digitizing ancient Japanese documents and art.

Distilling knowledge from Neural Networks to build smaller and faster models

A deep dive on distilling language models.

Introducing the Next Generation of On-Device Vision Models: MobileNetV3 and MobileNetEdgeTPU

Google releases checkpoints for MobileNetV3, showing 15-30% speedups and increased accuracy compared to previous versions.

[GitHub] Kayzaks/HackingNeuralNetworks

A small course on exploiting and defending neural networks.

[GitHub] SachsLab/IntracranialNeurophysDL

A short course for getting started with deep learning for intracranial extracellular neurophysiology.

Libraries & Code

[GitHub] NVIDIAGameWorks/kaolin

PyTorch library aimed at accelerating 3D deep learning research.

[GitHub] victorqribeiro/hntitlenator

A neural network to predict whether your HN post will get up votes by the title.

[GitHub] facebookresearch/DeepFovea

Neural Reconstruction for Foveated Rendering and Video Compression using Learned Statistics of Natural Videos.

[Updated] BodyPix: Real-time Person Segmentation in the Browser with TensorFlow.js

Human pose estimation and body part segmentation.

Papers & Publications

Location-aware Upsampling for Semantic Segmentation

Abstract: ….[W]e present a Location-aware Upsampling (LaU) that adaptively refines the interpolating coordinates with trainable offsets. Then, location-aware losses are established by encouraging pixels to move towards well-classified locations. An LaU is offset prediction coupled with interpolation, which is trained end-to-end to generate confidence score at each position from coarse to fine. Guided by location-aware losses, the new module can replace its plain counterpart (e.g., bilinear upsampling) in a plug-and-play manner to further boost the leading encoder-decoder approaches.

SimpleShot: Revisiting Nearest-Neighbor Classification for Few-Shot Learning

Abstract: Few-shot learners aim to recognize new object classes based on a small number of labeled training examples. To prevent overfitting, state-of-the-art few-shot learners use meta-learning on convolutional-network features and perform classification using a nearest-neighbor classifier. This paper studies the accuracy of nearest-neighbor baselines without meta-learning. Surprisingly, we find simple feature transformations suffice to obtain competitive few-shot learning accuracies. For example, we find that a nearest-neighbor classifier used in combination with mean-subtraction and L2-normalization outperforms prior results in three out of five settings on the miniImageNet dataset.