"Machine learning approach for low-dose CT imaging yields superior results" via Nature Machine Intelligence
https://m.phys.org/news/2019-06-machine-approach-low-dose-ct-imaging.html
https://m.phys.org/news/2019-06-machine-approach-low-dose-ct-imaging.html
phys.org
Machine learning approach for low-dose CT imaging yields superior results
Machine learning has the potential to vastly advance medical imaging, particularly computerized tomography (CT) scanning, by reducing radiation exposure and improving image quality.
Facebook's machine learning system MelNet generated this AI voice clone of Bill Gates. Others voice clones can be heard here: https://audio-samples.github.io under the heading “Selected Speakers.”
ICML 2019 | Google, ETH Zurich, MPI-IS, Cambridge & PROWLER.io Share Best Paper Honours
https://medium.com/syncedreview/icml-2019-google-eth-zurich-mpi-is-cambridge-prowler-io-share-best-paper-honours-4aeabd5c9fc8
https://medium.com/syncedreview/icml-2019-google-eth-zurich-mpi-is-cambridge-prowler-io-share-best-paper-honours-4aeabd5c9fc8
Medium
ICML 2019 | Google, ETH Zurich, MPI-IS, Cambridge & PROWLER.io Share Best Paper Honours
The 36th International Conference on Machine Learning (ICML) kicked off Monday in California. The ICML is one of the world’s two top…
Energy-Based Models applied to the detection of machine-generated text, as opposed to human-produced text.
Brought to you from FAIR.
https://arxiv.org/abs/1906.03351
Brought to you from FAIR.
https://arxiv.org/abs/1906.03351
arXiv.org
Real or Fake? Learning to Discriminate Machine from Human Generated Text
Energy-based models (EBMs), a.k.a. un-normalized models, have had recent successes in continuous spaces. However, they have not been successfully applied to model text sequences. While decreasing...
Are you really into object recognition, but you are sick of looking at 2D boxes and 2D masks? Let's play with 3D shapes!
We build on Mask R-CNN and extend it to infer 3D meshes. Given an input image, we detect all objects, infer their 2D instance boxes and masks as well as their 3D object shapes, all end-to-end! Naturally, we call our approach Mesh R-CNN :D
Paper: https://arxiv.org/abs/1906.02739
Joint work with Jitendra Malik and Justin Johnson
P.S. This task is really hard!
We build on Mask R-CNN and extend it to infer 3D meshes. Given an input image, we detect all objects, infer their 2D instance boxes and masks as well as their 3D object shapes, all end-to-end! Naturally, we call our approach Mesh R-CNN :D
Paper: https://arxiv.org/abs/1906.02739
Joint work with Jitendra Malik and Justin Johnson
P.S. This task is really hard!
arXiv.org
Mesh R-CNN
Rapid advances in 2D perception have led to systems that accurately detect objects in real-world images. However, these systems make predictions in 2D, ignoring the 3D structure of the world....
Here are the COMPLETE Lecture notes on Professor Andrew Ng's
Stanford Machine Learning Lecture: https://www.holehouse.org/mlclass/
Stanford Machine Learning Lecture: https://www.holehouse.org/mlclass/
Tons of presentations from Embedded Vision Summit 2019
https://www.embedded-vision.com/platinum-members/embedded-vision-alliance/embedded-vision-training/downloads/pages/may-2019-summit-slides
https://www.embedded-vision.com/platinum-members/embedded-vision-alliance/embedded-vision-training/downloads/pages/may-2019-summit-slides
Edge AI and Vision Alliance
May 2019 Embedded Vision Summit Replay
Keynotes “Making the Invisible Visible: Within Our Bodies, the World Around Us and Beyond,” a Keynote Presentation from the MIT Media Lab June
Recent paper from CVPR'19 paper about neural free-viewpoint rendering of human avatars without reconstructing geometry
https://www.profillic.com/paper/arxiv:1905.08776
https://www.profillic.com/paper/arxiv:1905.08776
Profillic
Profillic: AI research & source code to supercharge your projects
Explore state-of-the-art in machine learning, AI, and robotics research. Browse papers, source code, models, and more by topics and authors. Connect with researchers and engineers working on related problems in machine learning, deep learning, natural language…
Apparently, GANs are used to create fake profile pictures on LinkedIn for international industrial espionage.
https://apnews.com/bc2f19097a4c4fffaa00de6770b8a60d
https://apnews.com/bc2f19097a4c4fffaa00de6770b8a60d
AP News
Experts: Spy used AI-generated face to connect with targets
LONDON (AP) — Katie Jones sure seemed plugged into Washington's political scene. The 30-something redhead boasted a job at a top think tank and a who's-who network of pundits and experts, from the centrist Brookings Institution to the right-wing Heritage…
What is the fuss about TensorFuzz?
It is the fun automated software “testing” for neural networks,
adapting traditional coverage guided fuzzing techniques.
Run #TensorFuzz to take your test coverage to levels other methods cannot reach (e.g. activation coverage, not just class coverage)
Great work by Augustus Odena and @Ian Goodfellow
Join the #ICML2019 talk at 9:40am today. Grand Ballroom
Read at https://arxiv.org/pdf/1807.10875.pdf
Code: https://github.com/brain-research/tensorfuzz
It is the fun automated software “testing” for neural networks,
adapting traditional coverage guided fuzzing techniques.
Run #TensorFuzz to take your test coverage to levels other methods cannot reach (e.g. activation coverage, not just class coverage)
Great work by Augustus Odena and @Ian Goodfellow
Join the #ICML2019 talk at 9:40am today. Grand Ballroom
Read at https://arxiv.org/pdf/1807.10875.pdf
Code: https://github.com/brain-research/tensorfuzz
Shapes and Context:
In-the-wild Image Synthesis & Manipulation
https://www.cs.cmu.edu/~aayushb/OpenShapes/
In-the-wild Image Synthesis & Manipulation
https://www.cs.cmu.edu/~aayushb/OpenShapes/
Few-Shot Adversarial Learning of Realistic Neural Talking Head Models
paper — arxiv
https://arxiv.org/pdf/1905.08233.pdf
video — youtube
https://www.youtube.com/watch?v=p1b5aiTrGzY
paper — arxiv
https://arxiv.org/pdf/1905.08233.pdf
video — youtube
https://www.youtube.com/watch?v=p1b5aiTrGzY
YouTube
Few-Shot Adversarial Learning of Realistic Neural Talking Head Models
Statement regarding the purpose and effect of the technology
(NB: this statement reflects personal opinions of the authors and not of their organizations)
We believe that telepresence technologies in AR, VR and other media are to transform the world in the…
(NB: this statement reflects personal opinions of the authors and not of their organizations)
We believe that telepresence technologies in AR, VR and other media are to transform the world in the…
Best Deep Learning Research of 2019 So Far
https://medium.com/@ODSC/best-deep-learning-research-of-2019-so-far-7bea0ed22e38
https://medium.com/@ODSC/best-deep-learning-research-of-2019-so-far-7bea0ed22e38
Medium
Best Deep Learning Research of 2019 So Far
We’re just about finished with Q1 of 2019, and the research side of deep learning technology is forging ahead at a very good clip. I…
Monotonic Infinite Lookback Attention for Simultaneous Machine Translation
Arivazhagan et al.: https://arxiv.org/abs/1906.05218
#ArtificialIntelligence #DeepLearning #MachineLearning
Arivazhagan et al.: https://arxiv.org/abs/1906.05218
#ArtificialIntelligence #DeepLearning #MachineLearning
arXiv.org
Monotonic Infinite Lookback Attention for Simultaneous Machine Translation
Simultaneous machine translation begins to translate each source sentence
before the source speaker is finished speaking, with applications to live and
streaming scenarios. Simultaneous systems...
before the source speaker is finished speaking, with applications to live and
streaming scenarios. Simultaneous systems...
Beating the bookies with their own numbers - and how the online sports betting market is rigged
Kaunitz et al.: https://arxiv.org/abs/1710.02824
#Gambling #ComputerScience #Statistics
Kaunitz et al.: https://arxiv.org/abs/1710.02824
#Gambling #ComputerScience #Statistics
arXiv.org
Beating the bookies with their own numbers - and how the online...
The online sports gambling industry employs teams of data analysts to build forecast models that turn the odds at sports games in their favour. While several betting strategies have been proposed...
Stabilizing the Lottery Ticket Hypothesis
Frankle et al.: https://arxiv.org/abs/1903.01611
#ArtificialIntelligence #DeepLearning #NeuralNetworks
Frankle et al.: https://arxiv.org/abs/1903.01611
#ArtificialIntelligence #DeepLearning #NeuralNetworks
Table-Based Neural Units: Fully Quantizing Networks for Multiply-Free Inference
Covell et al.: https://arxiv.org/abs/1906.04798
#ArtificialIntelligence #DeepLearning #MachineLearning
Covell et al.: https://arxiv.org/abs/1906.04798
#ArtificialIntelligence #DeepLearning #MachineLearning
arXiv.org
Table-Based Neural Units: Fully Quantizing Networks for...
In this work, we propose to quantize all parts of standard classification
networks and replace the activation-weight--multiply step with a simple
table-based lookup. This approach results in...
networks and replace the activation-weight--multiply step with a simple
table-based lookup. This approach results in...
Understanding Generalization through Visualizations. arxiv.org/abs/1906.03291
arXiv.org
Understanding Generalization through Visualizations
The power of neural networks lies in their ability to generalize to unseen data, yet the underlying reasons for this phenomenon remain elusive. Numerous rigorous attempts have been made to explain...
Adaptive Nonparametric Variational Autoencoder. arxiv.org/abs/1906.03288