tgeorgy / Mgan
Licence: mit
Masking GAN - Image attribute mask generation
Stars: ✭ 120
Programming Languages
python
139335 projects - #7 most used programming language
Labels
Projects that are alternatives of or similar to Mgan
Neural Doodle
Turn your two-bit doodles into fine artworks with deep neural networks, generate seamless textures from photos, transfer style from one image to another, perform example-based upscaling, but wait... there's more! (An implementation of Semantic Style Transfer.)
Stars: ✭ 9,680 (+7966.67%)
Mutual labels: deep-neural-networks, image-processing, image-manipulation
Starnet
StarNet
Stars: ✭ 141 (+17.5%)
Mutual labels: deep-neural-networks, image-processing, image-manipulation
Nimp
Nimp - Node-based image manipulation program.
Stars: ✭ 45 (-62.5%)
Mutual labels: image-processing, image-manipulation
Popbot
Color splash effects using Deep Learning
Stars: ✭ 61 (-49.17%)
Mutual labels: image-processing, image-manipulation
Imageviewer
HDR, PFM, DDS, KTX, EXR, PNG, JPG, BMP image viewer and manipulator
Stars: ✭ 71 (-40.83%)
Mutual labels: image-processing, image-manipulation
Photon
⚡ Rust/WebAssembly image processing library
Stars: ✭ 963 (+702.5%)
Mutual labels: image-processing, image-manipulation
Imagene
A General Purpose Image Manipulation Tool
Stars: ✭ 36 (-70%)
Mutual labels: image-processing, image-manipulation
Pix2pix
Image-to-image translation with conditional adversarial nets
Stars: ✭ 8,765 (+7204.17%)
Mutual labels: generative-adversarial-network, image-manipulation
Cometa
Super fast, on-demand and on-the-fly, image processing.
Stars: ✭ 8 (-93.33%)
Mutual labels: image-processing, image-manipulation
Lggan
[CVPR 2020] Local Class-Specific and Global Image-Level Generative Adversarial Networks for Semantic-Guided Scene Generation
Stars: ✭ 97 (-19.17%)
Mutual labels: generative-adversarial-network, image-manipulation
Cfsrcnn
Coarse-to-Fine CNN for Image Super-Resolution (IEEE Transactions on Multimedia,2020)
Stars: ✭ 84 (-30%)
Mutual labels: deep-neural-networks, image-processing
Faceaging By Cyclegan
Stars: ✭ 105 (-12.5%)
Mutual labels: deep-neural-networks, generative-adversarial-network
Tensorflow object counting api
🚀 The TensorFlow Object Counting API is an open source framework built on top of TensorFlow and Keras that makes it easy to develop object counting systems!
Stars: ✭ 956 (+696.67%)
Mutual labels: deep-neural-networks, image-processing
Image To Image Papers
🦓<->🦒 🌃<->🌆 A collection of image to image papers with code (constantly updating)
Stars: ✭ 949 (+690.83%)
Mutual labels: generative-adversarial-network, image-manipulation
Bringing Old Photos Back To Life
Bringing Old Photo Back to Life (CVPR 2020 oral)
Stars: ✭ 9,525 (+7837.5%)
Mutual labels: generative-adversarial-network, image-manipulation
St Cgan
Dataset and Code for our CVPR'18 paper ST-CGAN: "Stacked Conditional Generative Adversarial Networks for Jointly Learning Shadow Detection and Shadow Removal"
Stars: ✭ 13 (-89.17%)
Mutual labels: generative-adversarial-network, image-processing
Dna Gan
DNA-GAN: Learning Disentangled Representations from Multi-Attribute Images
Stars: ✭ 65 (-45.83%)
Mutual labels: generative-adversarial-network, image-manipulation
Contrastive Unpaired Translation
Contrastive unpaired image-to-image translation, faster and lighter training than cyclegan (ECCV 2020, in PyTorch)
Stars: ✭ 822 (+585%)
Mutual labels: generative-adversarial-network, image-manipulation
Concise Ipython Notebooks For Deep Learning
Ipython Notebooks for solving problems like classification, segmentation, generation using latest Deep learning algorithms on different publicly available text and image data-sets.
Stars: ✭ 23 (-80.83%)
Mutual labels: deep-neural-networks, image-processing
X Ray Classification
X-ray Images (Chest images) analysis and anomaly detection using Transfer learning with inception v2
Stars: ✭ 83 (-30.83%)
Mutual labels: deep-neural-networks, image-processing
Masking GAN - Generating image attribute mask (pytorch)
!Disclaimer
This project is pretty much abandoned. I have achieved following results after spending a lot of time tuning the model. Unfortunatelly I was not able to reproduce results, which means I was lucky to get a good random seed once. Also, I am not interested in research in this direction right now, so the problem won't be fixed any time soon. C'est la vie.
Motivation
When I first approached semantic manipulation problem there was no solution like CycleGAN, or later findings. And even now all of them produce artifacts.
Approach
- Use generator architecture with built-in segmentation.
- Mix original image with new patches through the segmentation mask.
- Train the whole network end-to-end.
- Use L1 identity loss to constrain Generator and reduce changes.
Instructions
I am using CelebA dataset to train the model. There are two files you would need to reproduce results: img_align_celeba.zip and list_attr_celeba.txt
You can download them from here http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html and put into {PROJECT_DIR}/data
After that initialize data and train the model by running
sh init_data.sh
python train.py
Results
- In case a person is already smiling it doesn’t do any changes at all.
- It handles poorly some extreme head angles.
- There are still artifacts.
Advices
Consider following advices if you want to build this kind of a model:
- Make sure your GAN model converges without appling mask and L1 loss.
Acknowledgments
The code is inspired by pytorch-CycleGAN-and-pix2pix . This paper GANimation: Anatomically-aware Facial Animation from a Single Image arXiv:1807.09251 describes similar training scheme.
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].