AIAIAI

Nvidia has released the NeRF model for public access. It's a neural network that creates videos and even 3D scenes from just a couple of photos. Essentially, it can do anything: create a clip, a cutscene for a game, and even plan a renovation. • GitHub _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

Animal-shaped marmalade. Midjourney. 🤖 Midjourney Telegram Bot _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

🧪 Supermarket AI-bot has learned to create recipes with poisons. In late July, New Zealand supermarket Pak’nSave launched a chat-bot named Savey Meal, based on AI to create recipes. However, it began producing unexpected results when users started indicating various household chemicals as ingredients. This was noticed by political commentator Liam Hehir, who suggested his subscribers on X (formerly Twitter) to experiment with the bot. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

😜 What happened next? The story of the famous meme generated by GEN-2. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

💔 New life hacks on TikTok: How to use neural networks to provoke jealousy in guys. In TikTok, girls are actively sharing life hacks on how to use Photoshop's neural network to make a guy jealous. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

Gladiators' Battle: Elon Musk has reached an agreement with the current leadership of Italy regarding the location of the duel with Mark Zuckerberg. According to the owner of X (formerly Twitter), "ancient Rome" will appear in the frame. The broadcast of the battle between the American entrepreneurs will be carried out on their social networks. The date of the fight has not yet been scheduled. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

📹 The Prodigy - Breathe Ai Video - Gen-2 by Runway. • YouTube (Note: The given text appears to be in English already and does not need translation.) _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

🗣 Freddiemeter — a site based on AI, developed by Google in collaboration with YouTube! Visit the site > grant "access to the microphone" (developers assure that data is not stored on servers) > follow the instructions. (For a microphone, it is best to use headphones (the device's microphone doesn't transmit sound as well)) In the end, AI will rate you from 0 to 100, and also check: >Pitch — how well you hit the notes. >Melody — how melodiously you were able to sing. >Timbre — how much your vocal style matches Freddie Mercury. • Try it _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

🤨 A girl, using the service of ElevenLabs, cloned her boyfriend's voice to test his fidelity. Having previously recorded all the important questions, she found out from a friend of her now ex-boyfriend that he had cheated on her. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

🪄 CapCut: Video Enhancement. A new service has emerged from CapCut for video enhancement. The newcomer has integrated AI, which can increase the resolution of clips in just a few seconds. 10 minutes per month are available for free. • Try it out _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

Do you consider images generated by artificial intelligence to be a legitimate form of creativity?🤔 At the Museum of Modern Art in New York MoMA, they definitely think so. This 3D animation is generated in real time. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

And it all started like this: _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

📹 An enthusiast shot a clip using neural networks for the Baddadan track by Chase & Status. It only took three hours to complete the work. First, he generated 200 frames in Midjourney, then animated them in Gen-2. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

Realistic Retro-Pokémon in Studio Ghibli style. Midjourney. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

Another fantastic cheat sheet on various styles and techniques for generating images based on SDXL. https://rikkar69.github.io/SDXL-artist-study/ _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

🍔 Advertising Spot Mcdonald's Delicious and that's it, generated by artificial intelligence using Gen-2. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

👾 This is what we play! :) DOOM has been launched on a keyboard button - the game was ported to a tiny keycap with an integrated microcontroller. However, there's a catch - to play on such a keyboard, you will need another keyboard. _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

What will be the capabilities of digital avatars in the future? There are about 850 muscles in the human body. In this video, researchers demonstrated that they can already simulate at least 90 of them in three-dimensional space. And this is without any simplifications and tricks currently used in 3D animation and motion capture technology. So, what will we see in a year or two? We are looking forward. Behind this work and more details at this link: sites.google.com/... _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

👾 The guy in the video is fully generated by a neural network. Voice, facial expressions, gestures - all this is AI. According to the developers, it took 2 minutes to generate the avatar. • Give it a try _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.

Oh, the Chinese have taken to generating avatars based on text and pose. There, they have pose control, ControlNet, and Nerfs. Basically, it’s a medley of known solutions for creating 3D avatars. They even claim that the resulting meshes are animatable: the generated avatars are easily animatable through skeletal binding, as they exhibit high alignment with the joints of the SMPL model. They insist on the high quality of their meshes. No code, only paper. https://avatarverse3d.github.io/ _______ This post was made possible by the voluntary work of the Stanford AI Research Lab.