Artificial Intelligence thread

T

tokenanalyst

Brigadier
Registered Member

Huawei releases new generation Hongmeng cockpit: equipped with Qianwu large model!​


At today's Huawei smart car solution conference, Huawei brought us the eye-catching new generation of Hongmeng cockpit, which integrates smart car machines, smart audio and smart displays, bringing an unprecedented smart experience to car owners.

This new generation of Hongmeng cockpit not only has the advanced Qianwu large model, but also integrates the Pangu large model, MindSpore Yisi computing framework and Yiteng AI basic hardware platform, jointly building a powerful 200Apps car and machine ecosystem.The Qianwu large model is a highlight of the Hongmeng cockpit. Its Qianwu vision function achieves millimeter-level accurate perception in the cabin. It can adaptively adjust the rearview mirror, steering wheel, HUD height, etc. according to the driver's habits, making driving more personalized.

In addition, Qianwu Vision also supports full-cabin skeletal-level human perception and multi-modal fusion car control technology. Car owners can easily control sunshades, doors, air-conditioning wind direction, etc. with just a wave of their hands, making driving operations more convenient.Not only that, the Hongmeng cockpit is also connected to the Xiaoyi voice assistant, which supports direct communication and cross-device scheduling, and provides personalized recommendations and suggestions, making information acquisition and interaction during driving more efficient.

Please, Log in or Register to view URLs content!
 
G

GulfLander

Just Hatched
Registered Member
Chinese team unveils first text-to-video AI model on par with Sora -By Global Times


Screenshot_20240427_184818_Chrome.jpg
Chinese tech-firm ShengShu-AI and Tsinghua University on Saturday unveiled text-to-video artificial intelligence (AI) model Vidu, which is said to be the first in China that's on par with Sora, in another manifestation of China's rapid development in the emerging critical AI field.

Launched at the ongoing Zhongguancun Forum in Beijing, Vidu can generate a 16-second 1080P videoclip with one click. It is built on a self-developed visual transformation model architecture called Universal Vision Transformer (U-ViT) integrating two text-to-video AI models of the Diffusion and the Transformer, the developers said.

The AI text-to-video model came just about two months after Sora, which is developed by the US-based developer OpenAI, was released to great fanfare worldwide.

"After the release of Sora, we found that it closely aligned with our technical roadmap, which further motivated us to advance our research with determination," Zhu Jun, vice dean of the Institute for Artificial Intelligence at Tsinghua University and chief scientist of ShengShu-AI, said at the forum.

The core technology of U-ViT was firstly proposed by Vidu's research team in September 2022, earlier than Sora's model architecture of DiT - Diversity in Transformation, which is the world's first visual transformation model architecture combining the advantages of the Diffusion and the Transformer, according to media reports.

During a live demonstration on Saturday, Vidu can simulate the real physical world and generate scenes with complex details in line with real physical laws, such as reasonable light and shadow effects and delicate facial expressions. It can also generate complex dynamic shots, instead of fixed ones.

Moreover, developed in China, Vidu has a great understanding of Chinese factors and can generate images of unique Chinese characters such as panda and loong, according to media reports. Ctto

Please, Log in or Register to view URLs content!
 
supercat

supercat

Major
A team of Chinese scientists says it has developed an artificial intelligence system to help doctors identify early-stage esophageal cancer, with results that double the detection rate of a naked-eye assessment by a doctor.
Click to expand...

Chinese AI innovation identifies oesophageal cancer lesions at double rate of doctors’ assessment by eye: paper​

  • Algorithm that highlights lesions on the gullet for doctors during an endoscopy is now used in hundreds of hospitals in China
  • Early detection is key as the cancer has five-year survival rate of more than 90 per cent when treated endoscopically or surgically before symptoms show
Click to expand...
Please, Log in or Register to view URLs content!
 
tphuang

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
tphuang said:
Alibaba cloud is first in China to offer Llama 3 training and inference
Please, Log in or Register to view URLs content!

Sensetime releases SenseNova 5.0, aimed at GPT4
Please, Log in or Register to view URLs content!
Claims a bunch of rankings doing better than GPT4

has 12000 PFLOPS computation now and will have 18000 PFLOPS by end of year with 2000 PLFOPS using domestic chips
Click to expand...
more details on Sensetime and SenseNova 5.0

To me, this seems like China's most complete LLM so far. Pretty impressive across the board

https://twitter.com/i/web/status/1784337819357401310
 
tphuang

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
China mobile opening up new AI Datacenter in Hohhot with 20000+ card and 6.7 EFLOPS of computation (apparently 85% using domestic chips). And based on my computation, it's mostly Ascend-910B

It will open up 2 more AI Datacenters this year of 20000 card each. So, a lot more Ascend chips getting added in.
 
You must log in or register to reply here.
Top