公开(公告)号：US20220253747A1

公开(公告)日：2022-08-11

申请号：US17616494

申请日：2020-05-26

Applicant: Google LLC

Inventor： Jie Ren ,  Balaji Lakshminarayanan ,  Peter Junteng Liu ,  Joshua Vincent Dillon ,  Roland Jasper Snoek ,  Ryan Poplin ,  Mark Andrew DePristo ,  Emily Amanda Fertig

IPC: G06N20/00 ,  G06N3/12

Abstract: The present disclosure is directed to systems and method to perform improved detection of out-of-distribution (OOD) inputs. In particular, current deep generative model-based approaches for OOD detection are significantly negatively affected by and struggle to distinguish population level background statistics from semantic content relevant to the in-distribution examples. In fact, such approaches have even been experimentally observed to assign higher likelihood to OOD inputs, which is opposite to the desired behavior. To resolve this problem, the present disclosure proposes a likelihood ratio method for deep generative models which effectively corrects for these confounding background statistics.

公开(公告)号：US20240169272A1

公开(公告)日：2024-05-23

申请号：US18551847

申请日：2022-02-07

Applicant: Google LLC

Inventor： Patricia MacWilliams ,  Abhijit Guha Roy ,  Jim Winkens ,  Alan Karthikesalingam ,  Jie Ren ,  Balaji Lakshminarayanan

IPC: G06N20/00

CPC classification number: G06N20/00

Abstract: A method includes determining, by a machine learning model and based on input data, a feature map that represents learned features present in the input data. The method also includes, for each respective inlier class of a plurality of inlier classes, determining, by the machine learning model and based on the feature map, a corresponding inlier score indicative of a probability that the input data belongs to the respective inlier class. The method additionally includes, for each respective outlier class of a plurality of outlier classes, determining, by the machine learning model and based on the feature map, a corresponding outlier score indicative of a probability that the input data belongs to the respective outlier class. The method further includes determining, based on the inlier scores and the outlier scores, whether the input data corresponds to the plurality of inlier classes or to the plurality of outlier classes.

公开(公告)号：US20240282131A1

公开(公告)日：2024-08-22

申请号：US18421672

申请日：2024-01-24

Applicant: Google LLC

Inventor： Jie Ren ,  Zhe Liu ,  James Urquhart Allingham ,  Michael Ward Dusenberry ,  Dustin Tran ,  Yin Cui ,  Balaji Lakshminarayanan ,  Xiuye Gu

IPC: G06V20/70 ,  G06F40/40 ,  G06V10/74 ,  G06V10/764 ,  G06V10/776

CPC classification number: G06V20/70 ,  G06F40/40 ,  G06V10/761 ,  G06V10/764 ,  G06V10/776

Abstract: Systems and methods for zero-shot prompt ensembling for zero-shot classification with text-image models can include utilizing a pre-trained text-image model to perform downstream tasks based on prompt-based weighting. The systems and methods may adjust for frequency-based bias and may automatically determine different prompt associations with a given downstream task. The systems and methods can aggregate weighted text embeddings and then determine a classification output based on similarity measures between an image embedding and the aggregated weighted text embeddings.