Computer Vision News - January 2022

29 Most Read in 2021 Coding Workshop: Creating a multi- object tracking model using a pre-trained RNN - with code! Our Ioannis Valasakis demonstrates a pre- trained network that combines two tasks in a single network to improve the inference speed. Here, a simple baseline is presented (with code!) to address the problems and it remarkably outperforms the state-of-the-arts on the MOT challenge datasets at 30 FPS. We hope that this baseline has inspired you and helped you evaluate new ideas. Read It Here “We are actually going to change the world!” with Raquel Urtasun of Waabi Awesome Raquel Urtasun has recently been included in the Entrepreneur’s 100 Women of Impact in 2021, and she has been starring for years now in our own inspiring group of over 100 Women in Computer Vision . Her newly founded company Waabi just raised a $100 million Series A initial funding. Raquel is set to revolutionize the world as we know it and she told us how she and her team will do it. Read It Here One-click Segmentation of Multi-Modal Medical Images by RSIP Vision One-click segmentation is a semi-automated, AI-based segmentation and measurement tool for detecting selected regions of interest and their boundaries quickly and automatically. In medical imaging especially, this tool has many benefits: high accuracy, generalization, user impact and minimal effort. Watch the short video and see why this so simple technique by RSIP Vision is so powerful. Read It Here