This AI Creates Human Faces From Your Sketches!
Footage taken from the paper "DeepFaceDrawing: Deep Generation of Face Images from Sketches".

https://www.instagram.com/p/CE6-JdCAof8/?igshid=l5168it3r358

To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI

Tutorial 1: https://lnkd.in/gSpeKx7

Code1: https://lnkd.in/gYq_nku

Code 2: https://lnkd.in/ggSmHwB

To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI

Contact only if you are experienced and serious:

Hello , We are looking for a dedicated developer who can work for a month.
The project is for a pet care start up
Skills required: angular JS / react native
Deliverable: web application
Duration: 1-2 months
Number of hours per week: 30-35

Budget 50k for entire project.

If interested plz reach out to me on 7719882295 or email anand@anikatechnologies.com

Real time face recognition with Android + TensorFlow Lite

The impressive effect of having the state-of-the-art running on your hands

https://medium.com/@estebanuri/real-time-face-recognition-with-android-tensorflow-lite-14e9c6cc53a5

To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI

First Order Motion Model for Image Animation. Using Deep Learning to animate an image by retargeting motion in an original video.
https://github.com/AliaksandrSiarohin/first-order-model

To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI

Using Augmented reality and Virtual reality to revolutionise restaurant experience!!! https://www.linkedin.com/posts/asif-immanad_innovation-augumentedreality-artificialintelligence-activity-6718905930135687168-Cxhd

All you need to Master Data Science!!!
https://drive.google.com/drive/mobile/folders/1sYSp6XOAg7urL2n_9QIwL4FDKUJH3r8U


To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI

Advance Data science Projects

1) Identify your Digits Dataset : https://www.kaggle.com/c/digit-recognizer/data

2) Recommendation Engine : https://cseweb.ucsd.edu/~jmcauley/datasets.html

3) Visual QA : https://visualqa.org/download.html

4) Vox Celebrity : http://www.robots.ox.ac.uk/~vgg/data/voxceleb/

5) Breast cancer classification : https://www.kaggle.com/martinab/breast-cancer-classification-wisconsin-dataset

6) Traffic signals : http://benchmark.ini.rub.de/?section=gtsrb&subsection=dataset

7) Image caption generator : https://academictorrents.com/details/9dea07ba660a722ae1008c4c8afdd303b6f6e53b

To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI

DeepFaceLab 2.0 Quick96 Deepfake at 1 Million Iterations
Featuring Iron Man's Robert Downey Jr. as Elon Musk.

To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI

I tried transfer learning on the pre-trained network. Custom dataset which was acquired from instagram by the custom scraper based on selenium library was used and the searching tag was 'flower'. The network was pre-trained with FFHQ dataset that consists of human face dataset. The way training network that is already trained with other dataset is called 'transfer learning'. In the middle of training, the network can give interesting and bizarre images which seem like mixture of flowers and human faces.
⠀
Generating large images like 1024*1024 is very hard task even using a neural network. That's because there're too many possibilities. If considering only a 1ch, 1024*1024, 8bit image, there can be 256^1,048,576 possibilities. What a large number! Most of the space will be filled with completely unrecognizable, noisy images and only very tiny and sparse portion of the space will be filled with natural images. What the network does is mapping input Gaussian noises to another space where natural images are. This space is called 'latent space'. In this space, there might be natural face images too. Since facial images belong to natural images, if learned well on facial images, modifying weights of the network to make flower images is much easier than just making flower images from the bottom.

GitHub link:-
https://github.com/okdalto/InteractiveGAN

To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI

I tried transfer learning on the pre-trained network. Custom dataset which was acquired from instagram by the custom scraper based on selenium library was used and the searching tag was 'flower'. The network was pre-trained with FFHQ dataset that consists of human face dataset. The way training network that is already trained with other dataset is called 'transfer learning'. In the middle of training, the network can give interesting and bizarre images which seem like mixture of flowers and human faces.
⠀
GitHub link:-
https://github.com/okdalto/InteractiveGAN

To learn AI from basics:
YouTube: https://www.youtube.com/c/asifimmanad

Telegram: https://t.me/Artificial_intelligence_AI