2022-03-12 08:11:00 CAFIAC FIX 2022-01-20 09:47:25 @Merzmensch Mona is given a blood transfusion. https://t.co/gXz0Mt6BJb 2022-01-20 09:39:14 @BadourAlBahar Thanks to you and your team for such a wonderful neural network! 2022-01-18 11:06:49 HomeStylist v.1.0 Neural network for generating photorealistic detail-preserving human reposing results and virtual try-on. Special thanks to @BadourAlBahar Try it online: https://t.co/fDrZImgx2l https://t.co/5oJmJWzCj8 2022-01-17 16:40:49 New image upscaling tool by https://t.co/oxVckRDtW8 - 5 free images - x2, x4 upscale - jpeg decompression - color correction - face enchantment - scratch removal Coming Soon: colorization 10% discount with LETSIMPROVE promo code https://t.co/CnQtt5UrHr https://t.co/0I5fzFQlaj 2022-01-17 08:11:00 CAFIAC FIX 2022-01-11 08:11:00 CAFIAC FIX 2022-01-01 14:30:05 Text2Voxel v.1.0 Generation of a voxel object from a text description. The 3D object takes about 40 minutes to generate. It is possible to write a request in any language. Special thanks to @apeoffire Try it online: https://t.co/OByCEqFcCb https://t.co/nlrkpCq0rs 2021-12-27 08:20:00 CAFIAC FIX 2021-12-17 08:59:33 @eyaler Thanks for the advice. I will definitely fix it. 2021-12-14 05:56:47 @XuanLuo14 @SIGGRAPHAsia The result is great 2021-12-13 14:14:06 NeuralWardrobe v.1.0 Generation of clothes by text description using ruDALL-E. Special thanks to @eyaler Try it online: https://t.co/hfackcgvm7 https://t.co/uiCcO2Ys2T 2021-12-10 12:22:58 UPD 1.1 - Added new inpainting method (FGVC). It works much longer, but better. Now we have coherence between frames. P.S. old method (left side), new method (right side) @Shedletsky @jbhuang0604 https://t.co/kfspWmlkcL 2021-12-10 10:46:45 @_aidan_clark_ Again, every technology has positives and negatives (and this one is no exception). You see the possibility of cheating, but I see a great tool to clean up. For example GPT-3 (or RETRO) can also be misleading, but without such technologies I can't imagine further development. 2021-12-10 10:02:56 @Shedletsky I'll try to do that in a future version. The main problem is not inpainting the object, but in its segmentation. 2021-12-10 09:58:57 @jbhuang0604 Interesting idea. I'll try to do it this weekend. 2021-12-10 01:59:51 @_aidan_clark_ I agree with you in part, but censorship, like everything else, has many advantages and disadvantages. Perhaps the name is a bit scary, but first of all it can be a great tool for videographers. In the future, it will be easier to remove objects that happen to be in the frame. 2021-12-08 14:20:46 NeuroCensor v.1.0 Removing objects in the video by text description. The video shows deleting "Grizzly bear" and "Man in a jacket". Works fast and supports multiple languages. Try it online: https://t.co/eZaiWTQBGH https://t.co/Xh9JM9wPgI 2021-12-07 16:59:19 @Norod78 Looks cool but "Runtime is limited to 4 hours when using GPU." 2021-11-18 04:58:55 @InnoKean The original dataset contains about 150k images, so I can try both. 2021-11-17 14:51:42 Images taken from: https://t.co/73jLwZow1C https://t.co/eOOmjKACHt 2021-11-17 14:50:15 Experiment with StyleGAN2-ADA. Tuned FFHQ-1024 on fashion dataset (10k images) for 150 hours. https://t.co/MZqwgzPyEz https://t.co/4L9lbJ6SOg 2021-11-09 18:21:23 Too cute result https://t.co/IseSaXHtzd 2021-11-06 23:20:00 CAFIAC FIX 2021-11-01 19:20:00 CAFIAC FIX 2021-11-01 17:30:00 CAFIAC FIX 2021-08-21 07:38:31 @shotacon_chris wait for the update 2021-08-20 09:10:25 @fizzylights @dribnet 500 2021-08-19 12:48:55 @dribnet Hello from Moscow https://t.co/RYc2ATPrkE 2021-08-18 19:15:52 @GerritGoewey @dribnet Prompt only. What is written under the picture I wrote in Colab. 2021-08-18 09:02:33 Text2PixelArt Pixel art generator based on text description using the CLIP model. Various languages ​​are supported. Special thanks to @dribnet Try it online: https://t.co/OTZEivUpdA https://t.co/LlWYzvMAGH 2021-08-08 18:55:54 @porestar It is. The only difference from the original is the structure of the small details. Globally, it is a complete copy, generated as a standard StyleGAN latent vector. 2021-08-08 12:56:48 @porestar Earrings are a little secret of Mona Lisa. 2021-08-07 14:09:42 #MonaLisa 2.0 As time goes on, technologies are improving. Now new Gioconda is even more realistic than my first version. https://t.co/oivCrglUCF 2021-07-02 12:34:55 @matheusmat ENTAR remains in history. Maybe someday I'll come back to upgrade it, but now I'm working on https://t.co/CaNJds7hUP algorithms 2021-03-12 11:16:15 @refikanadol Wow. What is the name of the screen model? 2021-02-15 19:32:41 Natural Language YouTube Searcher This is project, which allows you to search for particular moments in any YouTube video using CLIP model. Special thanks to @haltakov Try it online: https://t.co/Dmbsw359Bg https://t.co/HpPnpfi2aT 2021-02-11 04:51:56 @matheusmat I will definitely fix it when I have free time. 2021-01-30 16:53:22 Text2Image v.3.0 This neural network utilizes CLIP (with CMA-ES optimizer) to steer BigGAN towards images that match a textual prompt in any language. Special thanks to @eyaler Try it online: https://t.co/iFRPXf2t0v https://t.co/pNYbmz99D6 2021-01-18 16:33:27 Text2Image v.2.0 This neural network utilizes CLIP to steer BigGAN towards images that match a textual prompt in any language. Attached video shows #MonaLisa Special thanks to @advadnoun Try it online: https://t.co/BjNdgOP0D2 https://t.co/FKn6v1uXlH 2021-01-12 18:10:04 Text2Image Neural network for generating images from text in any language using the SIREN and CLIP architecture and models. Special thanks to @advadnoun Try it online: https://t.co/LuWC1hPOZL https://t.co/ojc0SpfMwh 2021-01-06 10:19:34 Boundless Fantasy DALL·E Try interactive examples: https://t.co/2KOB9SJg0r https://t.co/1phRCa1ofQ 2021-01-03 17:21:57 Oh my God, what an atmospheric retrospective. https://t.co/UewUpnuxpE 2020-12-14 18:34:42 @Norod78 Looks amazing! 2020-11-22 15:04:17 Portrait Painter A neural network that stylizes photos and videos into portrait drawing or animation. GitHub: https://t.co/3NzIntFTHG Try it online (RUS): https://t.co/QwlzaYtyux https://t.co/ltH6LSsRUk 2020-10-27 18:50:59 @pbaylies It looks amazing 2020-10-18 09:32:58 @eazybreezy57 @spakhm @arram @araghougassian @Bosefina @IgorBrigadir @jackconte @cpaik @zebulgar @DStrachman @fnthawar @genzmafia @minal_hasan @ikirigin @jordancooper @andreasklinger @jmwind It looks impressive, but it's hard to draw conclusions without demo. I would be interested to test Blanc as soon as possible. In the meantime, I'm wondering how your system deals with faces turned to some side? 2020-10-09 09:10:27 How face recognition works. (RUS) In this video, I talk about how faces are recognized in photos, compare the 6 most popular opensource models. Colab at the end of the video. https://t.co/cFpEtKLlRR 2020-09-16 17:58:48 Young #MonaLisa #toonify https://t.co/wsz4ZPcvqp 2020-09-16 17:07:08 Amazing result. #MonaLisa https://t.co/QsgzC4IQrp https://t.co/5jObupIxdU 2020-09-02 15:15:00 @alexeev_eu It is not necessary to make a single Colab. Both models do similar things. Both Colabs are on my GitHub. You can first animate the photo (with FOMM) and then make it speak (with Wav2Lips). 2020-09-02 13:32:43 Wav2Lip In this work, investigate the problem of lip-syncing a talking face video of an arbitrary identity to match a target speech segment. https://t.co/xGNVNoFsiO https://t.co/2Ajcbjufwt 2020-09-02 06:33:08 @jlnhlfan Did you even open Colab? Everything is written in it. https://t.co/5DayXyn8cE 2020-09-01 19:20:14 @jlnhlfan The "Add shortcut to Drive" button should have added a folder with weights to your Google Drive (as shown in the screenshot). P.S. sorry for the Russian interface https://t.co/GNrxDESs2f 2020-09-01 19:14:04 @jlnhlfan So these are different accounts (where you save the weights and where you give access) 2020-09-01 19:11:10 @jlnhlfan Go to Colab and follow the instructions. It says that you need to save the weights on Google Drive, not on your desktop. https://t.co/ULJ5rxYcOM 2020-09-01 07:39:39 @jlnhlfan Right now I tested BabyGAN (strictly according to the instructions) and everything worked fine, thus making your comment useless. Check if you have a BabyGAN folder on your Drive? Soon I will make a tutorial where I will tell people without programming experience how to use this. 2020-08-29 16:22:57 Above are photos generated by #BabyGAN, and below are my real ones. https://t.co/DbWKGDq5iF 2020-08-26 17:54:11 @majegru "Clone Git repository and install all requirements": def generate_final_image() 2020-08-26 08:18:28 @The_TSAN You can publish it now. It would be interesting to see. I did this back in January, but put it on the back burner. When I found out that FaceApp also implemented this algorithm, I decided to publish my own. 2020-08-26 04:53:23 @majegru In fact, it is a global variable, I just forgot to declare it as such. 2020-08-25 23:22:47 #BabyGAN StyleGAN-based predictor of children's faces from photographs of theoretical parents. Also you can change: age, face position, emotions and etc. GitHub: https://t.co/Yddi977kg7 Try it online: https://t.co/EPxSOo2ltU https://t.co/O6QNIccnpb 2020-08-22 15:17:18 @Norod78 Thank. Fixed 2020-08-22 14:48:33 Young #MonaLisa Experiment 2 Colab: https://t.co/dir68yNqwW https://t.co/HVHk06JR7y 2020-08-21 16:06:38 @hardmaru 1x3 Subpixel Text https://t.co/iW6siTNtTJ 2020-08-19 05:08:28 @_Mortdani Fixed it 2020-08-18 22:20:59 Now in Colab Try it: https://t.co/CUhamzyLxW https://t.co/J6i8CeIrYr 2020-08-18 22:18:53 @BrunoxMC Try it: https://t.co/CUhamzyLxW