Engineers at the University of Washington have successfully demonstrated an insect sized robot that can fly wirelessly. The robot is powered by a laser pointed at it that it converts to electricity using an onboard circuit.
Google unveiled Duplex at this year's Google IO. Duplex is a mind-blowing AI voice bot that can have conversations with real people in order to complete tasks like dinner reservations.
The examples are amazing, the system can deal with interruptions and can clarify details during the call in a realistic manner.Have a listen to the call below:
Researchers at MIT have developed a self driving car system for country roads that does not rely on prior 3D maps. The MapLite system uses GPS and LIDAR to map the edge of the road and map the path of the road on the fly, without the need for curbs or road signs.
Microsoft have partnered with drone maker DJI to allow Microsoft Azure IoT Edge AI code to be deployed onto drones using an SDK. The video shows a demo of the partnership duing Microsoft Build 2018 where a DJI Mavic Air with an onboard image recognition AI detects a structural anomaly in some demo pipes.
Google have open-sourced Seurat, the tool they use to simplify VR scenes based on the viewers perspective, sort of a 3D to 2D conversion on the fly using millions of triangles. Triangles are the best. The code is available on GitHub.
Researchers have created a new version of CRISPR that can edit free floating DNA in a test-tube, instead of only being able to operate on DNA within a living cell. This CRISPR sandbox can be used to test CRISPR on specific test samples, which would allow custom CRISPR diagnosis of cancer cells. The team involved are looking for commercial partners to help deevlop a CRISPR-on-a-chip product for cancer diagnosis.
Researchers in Italy, on a project called In Codice Ratio, are using OCR and AI to begin to digitise hand written manuscripts from The Vatican Secret Archives. The challenge lies in creating an OCR system efficient enough in recognising hand written Latin script. The researchers have pioneered a mehtod called jigsaw segmentation, where a word is segmented into blocks analogous to pen strokes, as opposed to segmentation by letter.
Two recently published papers have demonstrated significant advances in neuromorphic computing. The brain on a chip systems use conductive plates separated by a switching medium in order to emulate the synapse structure of the brain, which allows data to be both stored and processed in situ.
According to Connie Chan, runners of the Shanghai marathon can use a facial recognition product to easily buy photos of themselves running the marathon, automatically sourced from every photo of the race.
Face recognition commercialization:
Researchers at Carnegie Mellon University and Disney Research have developed a relatively cheap system of creating a touch wall using electrodes and conductive paint called Wall++. The smart wall system can operate in two modes; the first capactive sensing mode to detect touch; and the second electromagnetic sensing mode to detect and track an EM device's location on the wall.(H/T to Mikael for sharing this)
