Toronto Creative Deep-Learning Group 🎹🎨🎲 cover photo

What we’re about

Hi everyone,

*Why this group?*

As we know, deep learning has taken over as the predominant machine learning tool to solve real-world problems with accuracies never witnessed before. Personally, I am always fascinated by this feat, as I was part of the era where neural networks were dead animals and also lucky to the see the rise of them as they go 'deeper' - thanks to the increased parallel computation power, massive data, and active role of tech giants like Google and Microsoft.

I personally feel this is the time, we actively pursue to bring neural networks to *bring out our creativities*. Several folks around the world have started it, check out cool projects like:

• https://quickdraw.withgoogle.com/,

• https://magenta.tensorflow.org/

• https://affinelayer.com/pixsrv/

I am very excited about the future where we can use neural networks to achieve the peak of creativity. Imagine a world, where we can go beyond the limitation of time and limited human cognitive capacity using these tools, and create masterpieces in music or arts or movie that allows us to express our creativity.

I believe the time is now. Let's start a conversation, right here in Toronto.

The group will NOT be limited to deep learning researchers but will be open to creative artists, or anyone who wants to explore how to create artistic contents using deep learning tools.

This is an experimental group, and I am always open to your suggestions!

- Some reference code for this meetup

https://github.com/debanga/Creative-AI-Meetup

------------------------------------------------------------------------------------------

*Briefly about myself*

It's Debanga. I'm a computer vision researcher and a Toronto-based entrepreneur.

I have traveled a lot across the world in the last few years for research and work. Travelling and entrepreneurship have changed the way I look at the people and the world. I have come far from the days when I used to stay inside the dorm room coding day and night, to understand how important the communication among fellow people outside your domain of work is - both for a more accepting mindset and to increase our happiness. It provided me with an opportunity to learn more about others.

Verbal communication is not the only way of effective communication, we can express our emotions and passions through a piece of music, a painting, or even through a cool html5 website you just designed. But, we are limited by time - both in terms of learning and creating. This group is my attempt to start a conversation about how we can use the latest deep learning tools to make our passions come to life with the limited resources we have.

-----------------------------------------------------------------------------------------------------

*References*:

- Introduction to GAN - the most recent DNN sensation for creating novel data!

https://skymind.ai/wiki/generative-adversarial-network-gan

- Toward Multimodal Image-to-Image Translation

https://papers.nips.cc/paper/6650-toward-multimodal-image-to-image-translation.pdf

- StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks

https://arxiv.org/abs/1612.03242

- PROGRESSIVE GROWING OF GANS FOR IMPROVED QUALITY, STABILITY, AND VARIATION

https://arxiv.org/pdf/1710.10196.pdf

- A Style-Aware Content Loss for Real-time HD Style Transfer

https://arxiv.org/abs/1807.10201