From Recognition to Cognition: Visual Commonsense Reasoning

Zellers et al.: https://lnkd.in/ez3R-yq

#ComputerVision #PatternRecognition #Reasoning #machinelearning #technology


❇️ @AI_Python
✴️ @AI_Python_EN

Open Sourcing Active Question Reformulation with Reinforcement Learning

https://ai.googleblog.com/2018/10/open-sourcing-active-question.html

Natural language understanding is a significant ongoing focus of Google’s AI research, with application to machine translation, syntactic and semantic parsing, and much more. Importantly, as conversational technology increasingly requires the ability to directly answer users’ questions, one of the most active areas of research we pursue is question answering (QA), a fundamental building block of human dialogue.

❇️ @AI_Python
✴️ @AI_Python_EN

Three EPIC machine learning, Python and NLP goodies I've been looking at this past week:

1. Chartify by Spotify

How you communicate your work is just as important as the work itself.

Spotify was unhappy with the plethora of tools used for making visualisation in Python so they made their own.

And boy have they created some beautiful graphs.

Repo:

http://bit.ly/2RoTVwF

2. DeepSpeech by Mozilla

I've got to shed some love for the Mozilla team. All their work is world class.

They weren't satisfied with all the best speed to text models being locked up a cloud providers server somewhere.

So they reached out all over the internet, made the second largest voice database on the planet, replicated Baidu, Inc.'s DeepSpeech model and open sourced the whole thing!

Check it here:

http://bit.ly/2RpCHiD

3. BERT by Google

BERT = Bidirectional Encoder Representation Transformers

❇️ @AI_Python
✴️ @AI_Python_EN

Let's just leave it as BERT.

Following in the tradition of giving state of the art hashtag#deeplearning libraries funky names, Google has changed the game of NLP with their latest model, BERT.

I was already a fan of Transformers (especially Bumblebee) but now I have another reason to love them more.

Step up your NLP:

http://bit.ly/2Q5Fb9o

❇️ @AI_Python
✴️ @AI_Python_EN

❇️ @AI_Python
✴️ @AI_Python_EN

The third prize for the Best Project Competition in our course Computer Vision for Faces went to Andy Liang.

He is a C# developer who built a real-time face swapping mobile app based on the Unity framework for this final project.

The results are neat!

#ComputerVision, #MachineLearning, and #AI are the skills that are very applications-oriented.

You do not need the backing of a large company to build your own product.

If you want to learn how to build stunning applications using cutting-edge computer vision and machine learning algorithms, you must consider our course.

🌎 https://lnkd.in/gx_z6jf


❇️ @AI_Python
✴️ @AI_Python_EN

Maths behind a generative adversarial network (GAN) model and why it is hard to be trained. Wasserstein GAN is intended to improve GANs’ training

✅ http://bit.ly/2LFvIym #AI

#MachineLearning #DeepLearning #DataScience

❇️ @AI_Python
✴️ @AI_Python_EN

Deep Learning course: lecture slides and lab notebooks

https://m2dsupsdlclass.github.io/lectures-labs/
#منابع
❇️ @AI_Python
✴️ @AI_Python_EN

The following is a list of free, open source books on machine learning, statistics, data-mining, etc.

#MachineLearning #DeepLearning #DataScience

🌎 Link Review

❇️ @AI_Python
✴️ @AI_Python_EN

This Google Tutorial (96 slides total) on Machine Learning is the best:

🌎 https://bit.ly/2kyUKne

#BigData #DataScience #NeuralNetworks #AI #DeepLearning #ML #Algorithms #DataScientists #ReinforcementLearning

❇️ @AI_Python
✴️ @AI_Python_EN

How Can You Build A Winning Machine Learning Algorithm?

http://bit.ly/2PV6tPv

#MachineLearning #ml #Algorithms

❇️ @AI_Python
✴️ @AI_Python_EN

Introducing Amazon Go and the world’s most advanced shopping technology

Amazon Go is a new kind of store featuring the world's most advanced shopping technology. No lines, no checkout – just grab and go! Watch the video

#AI #ArtificialIntelligence

Is this the future Of shopping?

❇️ @AI_Python
✴️ @AI_Python_EN

"Condensing paragraphs into sentences isn’t easy for artificial intelligence (AI). That’s because it requires a semantic understanding of the text that’s beyond the capabilities of most off-the-shelf natural language processing models. But it’s not impossible, as researchers at Microsoft recently demonstrated.

In a paper published on the preprint server Arxiv.org (“Structured Neural Summarization“), scientists at Microsoft Research in Cambridge, England describe an AI framework that can reason about relationships in “weakly structured” text, enabling it to outperform conventional NLP models on a range of text summarization tasks."

https://venturebeat.com/2018/11/06/microsoft-researchers-develop-ai-system-that-can-generate-articles-summaries/

❇️ @AI_Python
✴️ @AI_Python_EN

"Training Neural Nets on Larger Batches: Practical Tips for 1-GPU, Multi-GPU & Distributed setups"

By Thomas Wolf: https://lnkd.in/etyMzjQ

#ArtificialInteligence #DeepLearning #MachineLearning #NeuralNetworks

❇️ @AI_Python
🗣 @AI_Python_arXiv
✴️ @AI_Python_EN

Applying artificial intelligence for social good

🌎 http://bit.ly/2raCQLp

#AI #ArtificialIntelligence #MachineIntelligence #ML #MachineLearning #DataScience #Analytics
#BigData #IoT

❇️ @AI_Python
🗣 @AI_Python_arXiv
✴️ @AI_Python_EN

Playing first-person shooter games with webcam and #DeepLearning (Tensorflow #ObjectDetection)

Find out how you can use an object detection model to control and play any first-person shooter game with your computer's webcam. Links to the code below.

Full Video: https://lnkd.in/eBq7z4r

Blog: https://lnkd.in/eekrqWk

Code: https://lnkd.in/ekhwwiJ

Subscribe: youtube.com/c/DeepGamingAI

@AI_Python
🗣 @AI_Python_arXiv
✴️ @AI_Python_EN

WOAH! OpenAI just released Spinning Up. A FREE resource for learning Deep Reinforcement Learning (RL).

Why Deep RL?

Because the beautiful thing about #DeepRL is much of the learning takes place by an agent in a virtual environment.

But wait... what's an agent? And what's an environment?

There's a new treatment available for patients with a certain issue. But she's hesitant to try it. She wants to wait for more trials to take place and more evidence to come out.

But running such trials in the real world is expensive and potentially harmful.

What if you could create a computer-generated version of Jessica (the agent), to try the treatment in a simulated medical centre (the environment) to see how it affected patients with similar characteristics to those in the real world?

With the knowledge you gain in the generated world, you could potentially improve the treatment and better suit it to each individual patient. Click HERE

❇️ @AI_Python
✴️ @AI_Python_EN

CS231n: Convolutional Neural Networks for Visual Recognition ,This is the syllabus for the Spring 2018 iteration of the course.

Schedule and Syllabus university of stanford

http://cs231n.stanford.edu/syllabus.html

❇️ @AI_Python
🗣 @AI_Python_arXiv
✴️ @AI_Python_EN

Artificial Intelligence Projected Revenue

#artificialintelligence #ai

❇️ @AI_Python
🗣 @AI_Python_arXiv
✴️ @AI_Python_EN

Difference Between

#BusinessIntelligence and #DataScience

❇️ @AI_Python
🗣 @AI_Python_arXiv
✴️ @AI_Python_EN

🔥To our new subscribers 🔥

If you have trouble reading our posts in farsi which is 100% normal if you are not a native persian speaker 😃 i invite you to join english version of our channel:
✴️ @AI_Python_EN

PS: we have another channel called arXiv with a great AI bot that posts significant and recent articles submitted to arXiv on a daily basis:
🗣 @AI_Python_Arxiv

Thank you for joining our community of AI researchers and Python users.

Meysam Asgari on behalf of ai_python admins team.

❇️ @AI_Python