Replace Dynamic Routing algorithm with EM Routing

This project needs complete knowledge on Capsule Network and a prior experience would be helpful.

Attached are the papers for Dynamic Routing as well as EM Routing of Capsules for the Capsule Network.


Need to change the following code:

[login to view URL]

It is currently using Dynamic Routing. Need to change a small part of it keeping rest the same to use EM Routing.

Project needs to be completed by 10-April, 2019.


I need this done in few hours.

I already train a CNN with transfer learning. The training is done

I need you to look at my code extract the file from the training and with that create a classifier.

Again the Network is trained and is all done. Is the last part of creating the classification itself.

Deep Learning App that recognizes multiple digits including decimals


i would like to have an android app that can recognize multiple digits in an image representing a price that i crop out and output the result with 2 decimals. Also it should be able to recognize prices i crop out from a receipt. Example of price and receipt crop can be seen in the attachment.

Additionally, i want you to write all the implementation details like what model you trained on, what architecture you used, etc. Contact me only if you have time to finish this project asap. More details over chat.

Rephrase articles using AI.

We need to rephrase an article given to AI and outcome must be another article which has same knowledge but in different grammer / sentence structure .

Example :

Mark was born in 1994 and he was rich by 30 .

AI should understand and make its own version as :-

Born in 1994 , Mark earned fortune by age 30 .

The articles can of 400 to 600 words in English only .

Need a deep learning / Machine Learning expert for Mask RCNN (Keras implementation) optimization with Intel Open Vino Toolkit

We need a person who is really capable of working for this project on urgent basis. The task is to use our pretrained keras model and convert it to intel OpenVino compatible formate (your suggestions will also be accepted). We also need to check our models performance on OpenVino.

There is no need of training or test for our pretrained model, we only want to make it OpenVino compatible

Please bid if you have knowlegde/worked on intel OpenVino before. Your suggestions will be acceptable for the desired deep learning model.

Link to vino: [login to view URL]

Weather Forecast for a Solar Tracker using Deep Learning

I have a solar tracker device and i would like to predict its energy efficiency using deep learning. Maybe you can make a CNN model in Keras on ResNet architecture that can predict on a public weather dataset, when the solar tracker will capture most of the energy according to weather predictions for the next 10 years or so.

Maybe you have a novel idea that can help me do this project, so im open to discussions.

Before accepting you to work on this project, i would like to know what is your approach and how you plan to make it. Thanks.