#سورس_کد #مقاله

دو پروژه مرتبط، یکی برای تشخیص موقعیت سه بعدی چهره، در شرایط سخت، از جمله
extreme viewing conditions, including near profile views, occlusions, and low resolution.

و پروژه دیگر، برای مدل سازی سه بعدی چهره، برای data augmentation

که میتوانید از ماتریس موقعیت پروژه اول، برای افزونگی داده در پروژه دوم استفاده کنید.

Face-Pose-Net
Estimate 3D face pose (6DoF) or 11 parameters of 3x4 projection matrix by a Convolutional Neural Network
https://github.com/fengju514/Face-Pose-Net

face_specific_augm
Face Renderer to perform Domain (Face) Specific Data Augmentation
https://github.com/iacopomasi/face_specific_augm

مقالات مرتبط

FacePoseNet: Making a Case for Landmark-Free Face Alignment
https://arxiv.org/abs/1708.07517

Do We Really Need to Collect Millions of Faces for Effective Face Recognition?
http://www.openu.ac.il/home/hassner/projects/augmented_faces/Masietal2016really.pdf

#face #pose #data_augmentation

تصویر اصلی

مرتبط با:
https://news.1rj.ru/str/cvision/879

#books_channel📚📚
#python
#deep_learning
- #CNN
- #LSTM
- #Capsulenet
#deep_tools
- #keras
- #tensorflow
- #theano
#data_mining
- #slide
- #implementation
.
.
.
🇯‌🇴‌🇮‌🇳 ↯

@Machine_learn

#مقاله
شمارش افراد در جمعیت
Mask-aware networks for crowd counting
https://arxiv.org/pdf/1901.00039.pdf

مقالات و پیاده سازی های شمارش افراد در جمعیت:

Awesome-Crowd-Counting
https://github.com/gjy3035/Awesome-Crowd-Counting

#Code
#Tools
#Datasets
#Papers
#Leaderboard


پست مرتبط:
https://news.1rj.ru/str/cvision/887


#Crowd_Counting

کامنت گذاری خودکار کدها با شبکه های LSTM !

Automatically generating code comments directly from source code using an LSTM. Works with multiple languages. Would be useful in any IDE.

https://www.twosixlabs.com/automatically-generating-comments-for-arbitrary-source-code/

#خبر
پیش بینی افراد سرشناس از هوش مصنوعی در سال 2019

AI predictions for 2019 from Yann LeCun, Hilary Mason, Andrew Ng, and Rumman Chowdhury
https://venturebeat.com/2019/01/02/ai-predictions-for-2019-from-yann-lecun-hilary-mason-andrew-ng-and-rumman-chowdhury/

Can we compress the knowledge of a large dataset into a small number of synthetically generated images? Researchers at FAIR, MIT, and Berkeley investigate in their paper: http://bit.ly/2GvWnAy

🙏Thanks to: @vahidreza01

#InstaGAN Excels in Instance-Aware Image-To-Image Translation

https://medium.com/syncedreview/instagan-excels-in-instance-aware-image-to-image-translation-64fb7d0344ae

#مقاله

یک کار جدید Image to image Translation
https://news.1rj.ru/str/cvision/892

مقاله:
https://arxiv.org/pdf/1812.10889.pdf

کد:
https://github.com/sangwoomo/instagan

The paper #InstaGAN: Instance-Aware Image-to-Image Translation has been accepted by the respected International Conference on Learning Representations (#ICLR) 2019, which will take place this May in New Orleans, USA.


This new research is based on #CycleGAN, a GAN variant which can learn to translate images without paired training data to overcome the limitations of one-by-one pairing of #pix2pix in image translation. CycleGAN can automatically translate two given unordered image sets X and Y, but it cannot encode instance information in an image. CycleGAN results however are not ideal when translating images involving specific features of the target. The InstaGAN system overcomes this problem and combines instance information from multiple task targets.

کارها و مطالب مشابه و مرتبط:
https://news.1rj.ru/str/cvision/214
https://news.1rj.ru/str/cvision/870
https-://t.me/cvision/863

#Image_to_Image_Translation #GAN

#سورس_کد
#InstaGAN
این مقاله نقص های cycleGan را رفع کرده.


#PyTorch implementation of "InstaGAN: Instance-aware Image Translation" (ICLR 2019)

code:
https://github.com/sangwoomo/instagan

paper:
https://arxiv.org/pdf/1812.10889.pdf

blog post:
https://news.1rj.ru/str/cvision/892
بیشتر:
https://news.1rj.ru/str/cvision/893

#Image_to_Image_Translation #GAN

سری توئیت های اندرونگ

1/The rise of Software Engineering required inventing processes like version control, code review, agile, to help teams work effectively. The rise of AI & Machine Learning Engineering is now requiring new processes, like how we split train/dev/test, model zoos, etc.

2/I'm also seeing many AI teams use new processes that haven't been formalized or named yet, ranging from how we write product requirement docs to how we version data and ML pipelines. This is an exciting time for developing these ideas!

3/Have you seen an idea for organizing ML projects that you'd like to share with others? If so please reply to this tweet!

https://twitter.com/AndrewYNg/status/1080886439380869122

#سورس_کد
Keras implementation of Sketch RNN
https://github.com/eyalzk/sketch_rnn_keras

#keras #Seq2SeqVAE #LSTM

#آموزش استفاده از TPU گوگل کولب و خواندن و پردازش داده های تصویری با tf.data dataset تنسرفلو

Here is an end-to-end canonical sample for training a model on Cloud TPUs in Keras. It has full code for loading the data from scratch using tf.data .Dataset and also exporting the trained model to ML Engine for inference.

Colab notebook:
https://colab.research.google.com/github/GoogleCloudPlatform/training-data-analyst/blob/master/courses/fast-and-lean-data-science/01_MNIST_TPU_Keras.ipynb

#TPU #keras #colab

#سورس_کد

#Mozilla has released open source #speech recognition model & data. Word error rate 6.5%, which is close to human.

Project DeepSpeech is an open source Speech-To-Text engine, using a model trained by machine learning techniques, based on Baidu's Deep Speech research paper. Project DeepSpeech uses Google's TensorFlow project to make the implementation easier.

Data: https://voice.mozilla.org/data
400k recordings, 500 hours of speech.

Model: https://github.com/mozilla/DeepSpeech
TensorFlow implementation of Baidu's DeepSpeech architecture.

https://deepspeech.readthedocs.io/en/latest/
DeepSpeech’s code documentation!

مرتبط با:
https://news.1rj.ru/str/cvision/875
https://news.1rj.ru/str/cvision/850

#speech_recognition #Tensorflow

#آموزش

Entity Embeddings For Categorical Data in Keras

فرض کنید داده های ساختاریافته را میخواهید به کراس بدهید. برای متغیرهای categorical چه میکنید؟ احتمالا آن‌ها را one-hot می‌کنید. مثلا برای روزهای هفته یک بردار به طول 7 که یکی از خانه های آن یک و بقیه 0 است!
اما راهی که خیلی وقت ها ما را به جواب بهتر میرساند استفاده از لایه embedding برای کد کردن روزهای هفته مثلا در یک وکتور به طول سه dense خواهد بود.
جزئیات پیاده سازی این کار در Keras و توضیحات را میتوانید بخوانید:

https://github.com/mayanksatnalika/ipython/tree/master/embeddings%20project
https://medium.com/@satnalikamayank12/on-learning-embeddings-for-categorical-data-using-keras-165ff2773fc9

#keras #embedding #categorical