Generating images on iOS devices using GANs
summary
The talk explains how to run GAN-based image generation models on iOS by converting Python-trained models into Core ML format. It covers specific architectures (DCGAN, Pix2Pix, CycleGAN, U-GAT-IT), locating the Generator within a trained model, using public pretrained models (Google tutorials and GitHub), and practical tips to smoothly convert and integrate TensorFlow/PyTorch models into an iOS app.
AI-written summary · about
chapters (7)
- 0:00GANs for image generation
- 3:24Core ML model conversion
- 10:04DCGAN to Core ML
- 17:54AnimeGAN to Core ML
- 27:22Freezing TensorFlow graphs
- 35:57Displaying images in iOS
- 39:51Summary
related talks
47mUnmasking Generative Adversarial Networks
41mTensorFlow Neural Networks on iOS
01h 03mTraining dynamic ML models on iOS 15
24mA Journey through Generative AI's Frontier Python Libraries
40mDeep Learning on iOS
34mDemystifying ML in iOS
40mOverview of Generative Adversarial Networks (GANs)
25mGetting Started with Machine Learning in your iOS apps