Wow 😲 , Lyft just open sourced its autonomous driving dataset from its Level 5 self-driving fleet!

Download: https://level5.lyft.com/dataset/
___________________________________________
For reference, the Lyft Level 5 Dataset includes:
1) Over 55,000 human-labeled 3D annotated frames;
2) Data from 7 cameras and up to 3 lidars;
3) A drivable surface map; and,
4) An underlying HD spatial semantic map (including lanes, crosswalks, etc.)
___________________________________________

SentiMATE: Learning to play Chess through Natural Language Processing. arxiv.org/abs/1907.08321

2019 Google Scholar Metrics Released, CVPR Cracks the Top Ten

Estimates peg the total number of academic papers and other scholarly literature indexed on the Google Scholar at almost 400 million, making it the world’s largest such database. To compile the index Google surveys hundreds of journals and websites that meet its inclusion guidelines, along with leading conferences in Engineering & Computer Science.

In a blog post last Friday Google released its 2019 Scholar Metrics, designed to “provide an easy way for authors to quickly gauge the visibility and influence of recent articles in scholarly publications …to help authors as they consider where to publish their new research.” One of the top AI conferences — IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) ranked in the top 10 for the first time, up from 20th in 2018. The world’s most prominent scientific journals, Nature and Science, ranked first and third respectively.

https://medium.com/syncedreview/2019-google-scholar-metrics-released-cvpr-cracks-the-top-ten-905deebbf833

Compositional Deep Learning. arxiv.org/abs/1907.08292

Good free course on ray-tracing, can be useful for those who want s to work in 3d area
https://www.youtube.com/playlist?list=PLujxSBD-JXgnGmsn7gEyN28P1DnRZG7qi

An Actor-Critic-Attention Mechanism for Deep Reinforcement Learning in Multi-view Environments

Elaheh Barati and Xuewen Chen : https://arxiv.org/abs/1907.09466

#reinforcementlearning #neuralnetwork #neuralnetworks #deeplearning

Mohammed Elhoseiny :

ICCV19: I am glad that our "Creativity Inspired Zero-shot Learning " paper got accepted at ICCV (1/1). A non-final version is available on Arxiv, and the final version will be available soon that has more experiments done recently, thanks to the reviewer's suggestions. The code will be made available.
https://arxiv.org/abs/1904.01109

The paper looks at generative zero-shot learning from creativity perspective where a likable visual generation from language description of an unseen visual class is achieved by a careful deviation from seen classes to enable better discrimination from seen classes, but not too much deviation to disable transfer from seen classes. The learning signal we propose is orthogonal to existing approaches and we show that it can be integrated with multiple generative ZSL approaches to improve them.

The AUC curve qualitatively illustrates the key advantage of our loss, doubling the capability of [57] from 0.13 AUC to 0.27 AUC to distinguish between two very similar birds in a 200 classification ways. The classes that barely differ by the curved feather coming from the head of the Crested Auklet vs the Parakeet Auklet class. Experiments on seven benchmarks were also performed showing the value of the loss.

A First Look at Quantum Probability, Part 2

Blog by Math3ma: https://www.math3ma.com/blog/a-first-look-at-quantum-probability-part-2

#QuantumProbability

deeplearning.ai : The final course of the deeplearning.ai TensorFlow Specialization is almost here! Kudos to those of you who already finished the first three courses .

New deep-learning approach predicts protein structure from amino acid sequence

https://hms.harvard.edu/news/folding-revolution

Learning Better Simulation Methods for Partial Differential Equations
Blog by Stephan Hoyer : https://ai.googleblog.com/2019/07/learning-better-simulation-methods-for.html
#climatechange #ai4good #machinelearning #partialdifferentialequations

New neural-network rain forecasting based on satellite images

Video: https://youtu.be/9zd3VR-prYU

Paper: https://arxiv.org/abs/1905.09932

Service: https://yandex.com/weather/nowcast

Reconstructing Visual Experiences from Brain Activity Evoked by Natural Movies
Nishimoto et al.: https://www.cell.com/current-biology/fulltext/S0960-9822(11)00937-7
#Brain #NeuralActivity #ResearchPapers

This is just the best speech ever. The presentation, the humour and the honesty is amazing. Not only that. The message in itself is something we need to work harder on. Now maybe more then ever before considering AI and other technology’s coming into commercial use. What should a young person starting school today graduating in 15-20 years actually focus on? What will be important skills, knowledge and behaviour of the future?


https://www.ted.com/talks/ken_robinson_says_schools_kill_creativity

#ai #knowledge #school #thefuture #technology

A Full Hardware Guide to Deep Learning

By Tim Dettmers: https://timdettmers.com/2018/12/16/deep-learning-hardware-guide/

#DeepLearning #hardware #machinelearning https://t.iss.one/ArtificialIntelligenceArticles

TensorFlow Lite for Microcontrollers

https://github.com/tensorflow/tensorflow/tree/master/tensorflow/lite/experimental/micro

Baidu's recent paper: Hubless Nearest Neighbor Search

Hubless Nearest Neighbor Search, a new method for Bilingual Lexicon Induction, improves retrieval accuracy significantly. Empirical results show HNN outperforms NN, ISF and other state-of-the-art.

Github: https://github.com/baidu-research/HNN
Paper: https://github.com/baidu-research/HNN/blob/master/doc/HNN.pdf

#ACL2019 #NLP #NLU