The big tech’s antitrust woes from last month have spilled over to this month as well. While the lawsuits, Twitter hacks and antitrust probes have occupied most part of the tech news, there are a few exciting releases for the tech-enthusiasts. In this week’s edition, Analytics India Magazine brings you all the interesting updates that have happened recently.
Now Your Phone Can Detect Earthquakes
Smartphones are equipped with expertly crafted accelerometers that detect the various movements of the phones. Now, Google is using these highly sensitive accelerometers as earthquake detectors. The giant network Android phones can now be used as a pool of seismometers that can alert a locality in real time. Know more about this project here.
Make Way For India’s Own SpaceX
India has now officially burst into the private space launch scene thanks to Skyroot. According to reports, the company, which was founded by former ISRO rocket engineers, has successfully tested the upper stage engineer burn. Founded a couple of years ago, this rocket startup has raised $4.3 million to date and is currently focused on developing its very first launch vehicle, the “Vikram-I,” which is scheduled for a launch next year. Know more details here.
Sign up for your weekly dose of what's up in emerging technology.
IIT Roorkee Launches Online AI Programmes
Indian Institute of Technology (IIT) Roorkee has partnered with Coursera, to launch Advanced Certification courses in AI and data science.
“IIT Roorkee is one of India’s top and oldest engineering schools, and millions aspire to learn from this illustrious institution.”-Jeff Maggioncalda, CEO of Coursera
This partnership makes IIT Roorkee one of the top universities such as Yale and Stanford, to join forces with Coursera to offer most in demand courses.
Download our Mobile App
UK’s Immigration Algorithm Botch Up
The UK’s visa application process has come under scanner for implementing algorithms that are institutionally racist. According to BBC, the screening system took some information provided by visa applicants and automatically processed it,and gave a colour code to each person based on a “traffic light” system – green, amber, or red.
“This streaming tool took decades of institutionally racist practices, such as targeting particular nationalities for immigration raids, and turned them into software.”-Chai Patel, JCWI
As a response to these allegations, the UK Home Office has announced that they are scrapping the streamlining tool to facilitate fairer visa screening process.
Quantum Computing Gets More Accessible
On Thursday, Amazon Web Services announced the general availability of Amazon Braket, a fully managed AWS service for designing quantum algorithms. Using Amazon Braket, researchers and developers can effortlessly get started with the technology and build quantum algorithms, test them on quantum circuit simulators, and run them on different quantum hardware technologies. Customers can use Amazon Braket to run their quantum algorithms on their choice of quantum processors based on different technologies, including systems from D-Wave, IonQ, and Rigetti. Know more about Braket here.
Video Calling On iPhone Just Got Better
NeuralCam, the company which builds AI powered camera apps for Apple phones has come up with a new app NeuralCam Live. The difference between the quality of a typical webcam and a phone camera with NeuralCam software is— Deep Learning.
The video enhancement that the algorithms offer can only be processed by customised hardware such as the A13 bionic chips in iPhones. NeuralCam app enables real-time video denoising and brightening models in this category achieve levels of quality and performance that are much better than traditional algorithms and make features such as video low light enhancement or face smoothing possible. Download the app here.
Austrian Researchers Uncover Flaws In ML Research
The researchers at the Institute for Artificial Intelligence and Decision Support in Vienna analyzed over 3,000 models on Papers with Code and found the results to be inconsistent. As appeared first on VentureBeat, the researchers looked at 32,209 benchmark results across 2,298 data sets from 3,867 papers published between 2000 and June 2020. The ML community puts in incredible amounts of energy to make their models state of the art. However, considering the survey by the Austrian researchers, it looks like the benchmarking methodologies need to be revisited.