Google Inc.

Mayank Kandpal

Filip Pavetic

Bakhodir Ashirmatov

Roman Vorushin

Jakub Kúdela

Methods and systems for improving precision of content matching systems at a platform are provided herein. Candidate matches for a media item are obtained. Each of the candidate matches indicates a reference media item including content that corresponds to content of the media item. Similarity data associated with the media item and the reference media items of the candidate matches is provided as input to a machine learning model. A determination is made, based on outputs of the model, of a content category associated with the media item and whether content of the media item matches content of a respective reference media item of the candidate matches in view of the determined content category. If so, action is taken to prevent users from the platform from accessing the content of the media item.

Precision of content matching systems at a platform

Video content screening using a trained video screening model trained using self-supervised training includes automatically generating a training dataset by obtaining predicate screening data indicating a predicate temporal segment within a training video and a corresponding reference temporal segment within the reference video, obtaining candidate screening data for an extended temporal segment from the training video, wherein the extended temporal segment includes the predicate temporal segment and at least one frame from the training video adjacent to the predicate temporal segment, wherein the candidate screening data indicates a similarity between a screening frame from the reference video and a spatial portion of a candidate frame from the extended temporal segment, and, in response to a determination that a determined similarity between the candidate subframe including, in the automatically generated training dataset, training example data indicating the similarity between the candidate subframe and the screening frame.

Video screening using a machine learning video screening model trained using self-supervised training

Generating a training image for use in training a region-of-interest detector that is trained to detect regions-of-interest within images includes generating a closed geometric shape; filling the closed geometric shape with a filler to obtain a blob; overlaying the blob on an edge of an image to obtain the training image, where the image includes a region-of-interest and a background region, and where the edge separates the region-of-interest from the background region; and using the training image to train the region-of-interest detector to detect a boundary of the region-of-interest. An input to the region-of-interest detector in a training phase includes the training image and a first indication of coordinates of the region-of-interest in the training image. An output of the region-of-interest detector includes a second indication of an area of the training image and a probability of the area of the training image being the region-of-interest.

Growing community of inventors

Zurich, Switzerland

Mayank Kandpal