公开(公告)号：WO2022272311A1

公开(公告)日：2022-12-29

申请号：PCT/US2022/073173

申请日：2022-06-25

Applicant: QUALCOMM INCORPORATED

Inventor： BORSE, Shubhankar Mangesh ,  CAI, Hong ,  ZHANG, Yizhe ,  PORIKLI, Fatih Murat

IPC: G06N3/04 ,  G06N3/08 ,  G06F18/23 ,  G06F18/2431 ,  G06N3/045 ,  G06N3/082

Abstract: Certain aspects of the present disclosure provide techniques for training neural networks using hierarchical supervision. An example method generally includes training a neural network with a plurality of stages using a training data set and an initial number of classification clusters into which data in the training data set can be classified. A cluster-validation set performance metric is generated for each stage based on a reduced number of classification clusters relative to the initial number of classification clusters and a validation data set. A number of classification clusters to implement at each stage is selected based on the cluster-validation set performance metric and an angle selected relative to the cluster-validation set performance metric for a last stage of the neural network. The neural network is retrained based on the training data set and the selected number of classification clusters for each stage, and the trained neural network is deployed.

公开(公告)号：WO2022272298A1

公开(公告)日：2022-12-29

申请号：PCT/US2022/073141

申请日：2022-06-24

Applicant: QUALCOMM INCORPORATED

Inventor： CAI, Hong ,  MATAI, Janarbek ,  BORSE, Shubhankar Mangesh ,  ZHANG, Yizhe ,  ANSARI, Amin ,  PORIKLI, Fatih Murat

IPC: G06T7/50 ,  G06T7/11 ,  G06N3/045 ,  G06N3/08 ,  G06T2207/10004 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/30252

Abstract: Certain aspects of the present disclosure provide techniques for cross-task distillation. A depth map is generated by processing an input image using a first machine learning model, and a segmentation map is generated by processing the depth map using a second machine learning model. A segmentation loss is computed based on the segmentation map and a ground-truth segmentation map, and the first machine learning model is refined based on the segmentation loss.

公开(公告)号：WO2023091925A1

公开(公告)日：2023-05-25

申请号：PCT/US2022/079927

申请日：2022-11-16

Applicant: QUALCOMM INCORPORATED

Inventor： BORSE, Shubhankar Mangesh ,  PARK, Hyojin ,  CAI, Hong ,  DAS, Debasmit ,  GARREPALLI, Risheek ,  PORIKLI, Fatih Murat

IPC: G06V10/20 ,  G06V10/764 ,  G06V20/70

Abstract: Aspects of the present disclosure relate to a novel framework for integrating both semantic and instance contexts for panoptic segmentation. In one example aspect, a method for processing image data includes: processing semantic feature data and instance feature data with a panoptic encoding generator to generate a panoptic encoding; processing the panoptic encoding to generate a panoptic segmentation features; and generating the panoptic segmentation mask based on the panoptic segmentation features.

公开(公告)号：WO2022192449A1

公开(公告)日：2022-09-15

申请号：PCT/US2022/019620

申请日：2022-03-09

Applicant: QUALCOMM INCORPORATED

Inventor： ZHANG, Yizhe ,  BORSE, Shubhankar Mangesh ,  PORIKLI, Fatih Murat

IPC: G06N3/04 ,  G06N3/08 ,  G06V10/40 ,  G06V10/82 ,  G06V20/40 ,  G06V10/75

Abstract: A method for processing a video includes receiving a video as an input at a first layer of an artificial neural network (ANN). A first frame of the video is processed to generate a first label. Thereafter, the artificial neural network is updated based on the first label. The updating is performed while concurrently processing a second frame of the video. In doing so, the temporal inconsistency between labels is reduced.