Skip to main content

Tensorflow Instance Segmentation



We have just released the instance segmentation support for the Tensor Flow Object Detection API. We support a number of instance segmentation models similar to those discussed in the Mask R-CNN paper. For further details refer to our slides from the 2017 Coco + Places Workshop. Refer to the section on Running an Instance Segmentation Model for instructions on how to configure a model that predicts masks in addition to object bounding boxes.


Comments

Popular posts from this blog

Bay Area House Prices in 2018

I have been studying the house market for the last few months, trying to see whether it is a good time to buy a house in the San Francisco bay area right now or we should wait for another few months or a year. I like to share few thoughts and resolutions. No body can predict the future obviously. The most honest answer anyone could provide you with when asked is simply to say that they don't know, unless they are biased for some reason. For example, if you ask people who own a house, they would always think the house prices are going to go up. Well, because that is what makes them feel happier. If you ask the real estate agents about their thoughts they would always be like `it is the best time both to buy a house and also the best time to sell a house`. Then you would be like how is it possible to be both. It should be either this or that. Well it is because that is what makes them feel happy. They make money if people sell houses and they make money if people buy houses. ...

AI-powered Personal Food Assistant

Today’s AI assistants help users with their everyday questions and curiosity, their commutes, and their product purchases. Despite the wide range of services, a neglected aspect, which consumes a large part of our everyday life,  is food . According to American Time Use Survey, Americans spend on average 1.18 hours per day in eating and drinking activities. More important than the time spent on food, what we eat is central to our health. An unhealthy food habit is likely to result in overweight, undernourished body conditions, and increase the risk of the development of nutrition-related diseases. However, we have not seen much investment on personalized AI assistants for healthy food choices. Albeit the lack of AI assistant for food, there are still hundreds of apps focusing on various aspects of food industry: Apps such as Instagram and Pinterest encourage users to take food pictures and share them with friends. Furthermore, these apps increase user retention with socia...