From: Frank Sully | 6/24/2021 6:37:00 PM | | | | Nvidia AI could let you make video calls in your PJs without anyone knowing
The model could make videoconferencing marginally more bearable
 STORY BY Thomas MacaulayWriter at Neural by TNW — Thomas covers AI in all its iterations. Likes Werner Herzog films and Arsenal FC.
Nvidia has unveiled an AI model that converts a single 2D image of a person into a “talking head” video.
Known as Vid2Vid Cameo, the deep learning model is designed to improve the experience of videoconferencing.
If you’re running late for a call, you could roll out of bed in your pajamas and disheveled hair, upload a photo of you dressed to impress, and the AI will map your facial movements to the reference image — leaving the other attendees unaware of the chaos behind the camera. That could be a boon for the chronically unkempt, but you should probably test the technique before you turn up in your birthday suit.
The system can also adjust your talking head’s viewpoint to show you looking straight at the screen, when secretly your eyes are fixed on a TV in the background.
They sound like nifty features for those of us who dread video calls, but the most useful aspect of the model may be bandwidth reduction. Nvidia says the technique can cut the bandwidth needed for video conferences by up to 10x.
GANimated video
Vid2Vid Cameo is powered by generative adversarial networks (GANs), which produce the videos by pitting two neural networks against each other: a generator that tries to create realistic-looking samples, and a discriminator that attempts to work out whether they’re real or fake.
This enables the two networks to synthesize realistic videos from a single image of the user, which could be a real photo or a cartoon avatar. During the video call, the model will capture their real-time motion and apply it to the uploaded image.
The model was trained on a dataset of 180,000 talking-head videos, which taught the network to identify 20 key points that encode the location of features including the mouth, eyes, and nose.
These points are then extracted from the image uploaded by the user to generate a video that mimics their appearance.
The model will be available soon in the Nvidia Maxine SDK and Nvidia Video Code SDK, but you can already try a demo of it here. I gave it a go, and was pretty impressed by the system’s efforts — although I wouldn’t let it cut my hair. |
| NVIDIA Corporation (NVDA) | Stock Discussion ForumsShare | RecommendKeepReplyMark as Last Read |
|
From: Frank Sully | 6/30/2021 6:06:01 PM | | | | Graphcore brings new competition to Nvidia in latest MLPerf AI benchmarks
Graphcore’s score in BERT natural language training was the best score of any of the two-socket AMD systems submitted.
 By Tiernan Ray | June 30, 2021 -- 17:00 GMT (10:00 PDT)
google.com |
| NVIDIA Corporation (NVDA) | Stock Discussion ForumsShare | RecommendKeepReplyMark as Last Read |
|
From: Frank Sully | 7/1/2021 11:22:24 AM | | | | Nvidia price target raised to new Street high of $1,000 at BMO Capital Markets
published: July 1, 2021 at 9:18 a.m. ET
By Emily Bary
BMO Capital Markets analyst Ambrish Srivastava boosted his price target on shares of Nvidia Corp. (NVDA) to $1,000 from $750, with the new target being the highest among analysts tracked by FactSet who cover Nvidia's stock. Shares are up 1.4% in premarket trading Thursday. Srivastava is upbeat about the growing relevance of software to the Nvidia business model. "As we look further out to the company's data center business, we now see the business growing to a $32 billion business a few years out versus our prior expectation of $25 billion," he wrote in a note to clients. "We also think that as software starts to become a bigger portion of the business, [gross margins] will continue to trend up, and we now see GMs of 75% for the business versus our prior modeled 72%." Nvidia shares closed at $800.10 in Wednesday's session. The shares are up 53% so far this year as the S&P 500 SPX, 0.20% has increased 14%. |
| NVIDIA Corporation (NVDA) | Stock Discussion ForumsShare | RecommendKeepReplyMark as Last Read |
|
| |