Computer Vision News - February 2023

25 AI Spotlight News MusicLM: Generating Music From Text Let’s end this month’s AI Spotlight News with a musical note: why couldn’t we use AI to generate the kind of music that best fits our mood and desires? That’s what they did at Google Research , where a team of researchers is introducing MusicLM , a model generating high-fidelity music from text descriptions such as " a calming violin melody backed by a distorted guitar riff ",whynot? Followthe link for examples. They also publicly release MusicCaps, a dataset composed of 5.5k music-text pairs, with rich text descriptions provided by human experts. Read More A Robot Was Scheduled to Argue in Court, Then Came the Jail Threats AI does no always win and not all the stories have a happy end. A British gentleman planned to have a “ robot lawyer ” help a defendant fight a traffic ticket in court. The person challenging a speeding ticket would wear smart glasses that both record court proceedings and dictate responses into the defendant’s ear from a small speaker. The system relied on the same AI text generators everyone’s talking about. To cut a long story short, his idea met a strong opposition from the side of corporations feeling threatened by it. Some of them answered with threats of prosecution and prison time. That’s how the story ends this time... Read More Open-Source Active Learning Toolkit for Computer Vision What machine learning engineer and data scientist doesn’t want to understand and improve their training data quality and help boost model performance ? That’s the promise of an AI-assisted platform provider called Encord . They have released a free open-source industry agnostic all-in-one toolkit designed for that purpose. One of the motivations disclosed by the company is that self-driving cars and diagnostic medical models using AI suffer from a “ production gap ”: proof-of-concept models perform well in research environments but struggle to make predictions accurately and consistently in real- world scenarios. Read More