In the paper See the Glass Half Full: Reasoning about Liquid Containers, their Volume and Content, one of the inputs to the model is "a bounding box mask smoothed by a Gaussian kernel". I'm not sure I understood what a "bounding box mask" is and how it is derived given the 4 coordinates (x1, x2, y1, y2) of the object's region in the image. Based on Figure 3, is this simply done by creating a black rectangle of the same dimensions as the image, and marking the region denoted by the coordinates with white pixels and then applying Gaussian smoothing to the entire image?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.