Hello
Generative Adversarial Networks have been used in image and video generation, mostly to generate infinite random results with some art / creative properties.
Do they have any other application in engineering or technology? Perhaps in creating random training sets?
Will
Hello Will Serrano ,
there are several applications with styleGAN to generate photos of people based on https://thispersondoesnotexist.com
Wireframing: UI prototypes designed with tools like Adobe XD, Sketch or Figma can be produced with UI Faces plugin for Figma, Adobe XD, Figma https://uifaces.co/?provider[]=11 Uing data in UX prototype i not a new tendency a we can see:
- https://www.blended.io/blog/designing-with-live-data.html
- https://blog.adobe.com/en/2019/04/01/5-adobe-xd-plugins-for-designing-with-real-data.html
Personas: You can read this question where we talk about personas and styleGAN https://www.researchgate.net/post/Have_you_considered_using_StyleGAN_to_generate_realistic_faces_for_personas_based_on_data
Marketing: Other applications are free marketing assets to avoid to make footages or buy photos. In the past as a webdesigner I created people by mixing faces in Photohop by playing with layers.
Augmented Data: Of course you can create datasets to test automatic systems and generate a lot of new face types that we don't have in common datasets. For example, you can generate huge variations of racial, gender, age faces to check efficiency of a system to discriminate different types of people with less issues. Suggested solution to try to limit racism issues in dataset: https://medium.com/@albertogrivera/racism-in-machine-learning-cac34b73bd4b
Self-supervised learning: self-supervised learning is possible by rotating the entire input images. The model learns to predict which rotation is applied and you can use GAN to predict form or configuration of a picture like https://lilianweng.github.io/lil-log/2019/11/10/self-supervised-learning.html and of course Yann LeCun seminal research https://www.facebook.com/watch/?v=2301388736824154
Test Automation: It could be great for the kind of test authentification or security described in this article Article 3D Face Authentication Software Test Automation
Augmented Zoology: As GAN is good to generate animal faces too, it's a good tool to calibrate animal face (or any kind of thing) recognition from videos Article Chimpanzee face recognition from videos in the wild using de...
Will Serrano
A new kind of application for Facial Recognition Technology (FRT) which is able to differentiate between pigs by analysing their snouts, ears, and eyes. The system used in Guangxi farms constantly tracks pigs’ pulses and sweat rates; at the same time, voice recognition software monitors individual animals’ cough rates.
https://www.theguardian.com/environment/2020/dec/10/facial-recognition-for-pigs-is-it-helping-chinese-farmers-or-hurting-the-poorest
Facial Recognition Technology (FRT): Face detection technology determines whether the image contains a face. Face analysis technology aims to identify attributes such as gender, age, or emotion from detected faces. Face recognition technology compares an individual's facial features to available images for verification or identification purposes.
Aesthetics, Arts Sciences, history and archeology:
Enhancing Human Pose Estimation in Ancient Vase Paintings via Perceptually-grounded Style Transfer Learning
https://arxiv.org/abs/2012.05616
pdf: https://arxiv.org/pdf/2012.05616
Preprint:
https://www.researchgate.net/publication/346933628_Enhancing_Human_Pose_Estimation_in_Ancient_Vase_Paintings_via_Perceptually-grounded_Style_Transfer_Learning
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training