Top 3 AI Updates from Baidu’s Developer Conference 2018
- Baidu held it’s annual Developer conference in Beijing this week with a heavy focus on Artificial Intelligence
- Big announcements included China’s first fully autonomous bus, an AI chip and a huge upgrade to Baidu Brain
- Check out details on these three announcements in the article below!
Labelled as ‘China’s Google’, Baidu has made some really exciting announcements on AI-based innovations and product releases during their Create Developer conference in Beijing this week. We bring to you the top three AI focused announcements in this article.
China’s first Autonomous Bus
Baidu announced China’s first fully autonomous bus, Apolong, powered by the popular open-source Apollo autonomous driving platform. Apolong is equipped with Baidu’s voice-activated software and is able to perform driverless tasks (like obstacle avoidance or automatic transshipment). The first country to use this bus (outside of China) will be Japan.
Baidu also released an open source platform, Apollo, in July 2017. It now has more than 100 partners including Ford, Intel, NVIDIA, and Microsoft, among others. The team announced in the conference that this Apollo platform has been upgraded to Apollo 3.0, with improved support for autonomous driving in geo-fenced areas.
Since this is open-source, you can download the dataset that goes into making self-driving cars.
China’s first edge-to-cloud chip
Baidu unveiled a new chip, Kunlun, that can handle AI models for edge computing. Built to assist high performance requirements, specifically for AI purposes, it is able to run intensive computing operations on both device-based edge computing, and cloud-based through datacentres. The Kunlun 818-300 model is being built to train AI, and the 818-100 model has been designed for inference purposes. Baidu claims that Kunlun is approximately 30 times faster than their first FPGA (field-programmable gate array) chip. It can reach 260 tera-operations per second with 512 GBps memory bandwidth.
Baidu Brain 3.0
Baidu Brain was first introduced in September 2016 with the goal of expanding AI services and improving semantic understanding. Tt has now been updated to a new version, called Baidu Brain 3.0. This latest version features more than 110 AI services including, natural language processing, facial recognition, voice processing and recognition, among other things!
DuerOS 3.0 has also been announced along with Baidu Brain, which is Baidu’s platform for conversational AI. It works with smart devices and appliances like televisions, fridges, mobiles, etc.
Our take on this
The most attention garnering announcement was, you guessed it, the autonomous bus. Self-driving technology has been smack in the middle of the spotlight this year and Baidu has enhanced their name in the domain with this launch. Their popular open-source Apollo platform has always been a source of great use for the ML community and I encourage you to download their dataset and play around with it.
Which of the three announcements excited you the most? Let me know in the comments below!
Subscribe to AVBytes here to get regular data science, machine learning and AI updates in your inbox!