-
公开(公告)号：US11068741B2
公开(公告)日：2021-07-20
申请号：US16224644
申请日：2018-12-18
Applicant: QUALCOMM Incorporated
Abstract: Techniques and systems are provided for determining features for one or more objects in one or more video frames. For example, an image of an object, such as a face, can be received, and features of the object in the image can be identified. A size of the object can be determined based on the image, for example based on inter-eye distance of a face. Based on the size, either a high-resolution set of features or a low-resolution set of features is selected to compare to the features of the object. The object can be identified by matching the features of the object to matching features from the selected set of features.
-
公开(公告)号：US20190311183A1
公开(公告)日：2019-10-10
申请号：US15948676
申请日：2018-04-09
Applicant: QUALCOMM Incorporated
Inventor： Lei Wang , Yingyong Qi , Ning Bi
Abstract: Methods, systems, and devices for object recognition are described. A device may generate a subspace based at least in part on a set of representative feature vectors for an object. The device may obtain an array of pixels representing an image. The device may determine a probe feature vector for the image by applying a convolutional operation to the array of pixels. The device may create a reconstructed feature vector in the subspace based at least in part on the set of representative feature vectors and the probe feature vector. The device may compare the reconstructed feature vector and the probe feature vector and recognize the object in the image based at least in part on the comparison. For example, the described techniques may support pose invariant facial recognition or other such object recognition applications.
-
公开(公告)号：US10360456B2
公开(公告)日：2019-07-23
申请号：US15400118
申请日：2017-01-06
Applicant: QUALCOMM Incorporated
Abstract: Techniques and systems are provided for maintaining lost blob trackers for one or more video frames. In some examples, one or more blob trackers maintained for a sequence of video frames are identified. The one or more blob trackers are associated with one or more blobs of the sequence of video frames. A transition of a blob tracker from a first type of tracker to a lost tracker is detected at a first video frame. For example, the blob tracker can be transitioned from the first type of tracker to the lost tracker when a blob for which the blob tracker was associated with in a previous frame is not detected in the first video frame. A recovery duration is determined for the lost tracker at the first video frame. For one or more subsequent video frames obtained after the first video frame, the lost tracker is removed from the one or more blob trackers maintained for the sequence of video frames when a lost duration for the lost tracker is greater than the recovery duration. The blob tracker can be transitioned back to the first type of tracker if the lost tracker is associated with a blob in a subsequent video frame prior to expiration of the recovery duration. Trackers and associated blobs are output as identified blob tracker-blob pairs when the trackers are converted from new trackers to trackers of the first type.
-
公开(公告)号：US20190220653A1
公开(公告)日：2019-07-18
申请号：US15869342
申请日：2018-01-12
Applicant: QUALCOMM Incorporated
Inventor： Lei Wang , Ning Bi , Yingyong Qi
Abstract: Methods, systems, and devices for object recognition are described. Generally, the described techniques provide for a compact and efficient convolutional neural network (CNN) model for facial recognition. The proposed techniques relate to a light model with a set of layers of convolution and one fully connected layer for feature representation. A new building block of for each convolution layer is proposed. A maximum feature map (MFM) operation may be employed to reduce channels (e.g., by combining two or more channels via maximum feature selection within the channels). Depth-wise separable convolution may be employed for computation reduction (e.g., reduction of convolution computation). Batch normalization may be applied to normalize the output of the convolution layers and the fully connected layer (e.g., to prevent overfitting). The described techniques provide a compact and efficient CNN model which can be used for efficient and effective face recognition.
-
公开(公告)号：US10140718B2
公开(公告)日：2018-11-27
申请号：US15384802
申请日：2016-12-20
Applicant: QUALCOMM Incorporated
Inventor： Ying Chen , Lei Wang , Jinglun Gao , Ning Bi
Abstract: Techniques and systems are provided for processing video data. For example, techniques and systems are provided for maintaining blob trackers for one or more video frames. A blob tracker can be associated with a blob generated for a video frame. The blob includes pixels of at least a portion of one or more foreground objects in the video frame. The blob tracker can be determined to be a first type of tracker or a second type of tracker. A first type of tracker has a first bounding box and a second bounding box with an overlapping ratio greater than an alignment threshold for the first type of tracker. A second type of tracker has an irregular size change or an irregular motion change over a threshold duration. The blob tracker can be removed from the plurality of blob trackers maintained for the one or more video frames when the blob tracker is the first type of tracker or the second type of tracker.
-
Patent Agency Ranking
公开(公告)号：US10115005B2
公开(公告)日：2018-10-30
申请号：US15384997
申请日：2016-12-20
Applicant: QUALCOMM Incorporated
Inventor： Jinglun Gao , Ying Chen , Lei Wang , Ning Bi
Abstract: Techniques and systems are provided for processing video data. For example, techniques and systems are provided for performing context-aware object or blob tracker updates (e.g., by updating a motion model of a blob tracker). In some cases, to perform a context-aware blob tracker update, a blob tracker is associated with a first blob. The first blob includes pixels of at least a portion of one or more foreground objects in one or more video frames. A split of the first blob and a second blob in a current video frame can be detected, and a motion model of the blob tracker is reset in response to detecting the split of the first blob and the second blob. In some cases, a motion model of a blob tracker associated with a merged blob is updated to include a predicted location of the blob tracker in a next video frame. The motion model can be updated by using a previously predicted location of blob tracker as the predicted location of the blob tracker in the next video frame in response to the blob tracker being associated with the merged blob. The previously predicted location of the blob tracker can be determined using a blob location of a blob from a previous video frame.
-
公开(公告)号：US10019633B2
公开(公告)日：2018-07-10
申请号：US15384911
申请日：2016-12-20
Applicant: QUALCOMM Incorporated
Inventor： Ying Chen , Lei Wang , Jinglun Gao
CPC classification number: G06K9/00771 , G06K9/3233 , G06K9/6201 , G06K2009/6213 , G06T7/20 , G06T2207/10004 , G06T2207/10016
Abstract: Techniques and systems are provided for processing video data. For example, techniques and systems are provided for matching a plurality of bounding boxes to a plurality of trackers. In some examples, a first association is performed, in which case one or more of the plurality of bounding boxes are associated with one or more of the plurality of trackers by minimizing distances between the one or more bounding boxes and the one or more trackers. A set of unmatched trackers are identified from the plurality of trackers after the first association. The set of unmatched trackers are not associated with a bounding box from the plurality of bounding boxes during the first association. A second association is then performed, in which case each of the set of unmatched trackers is associated with an associated bounding box from the plurality of bounding boxes that is within a first pre-determined distance. A set of unmatched bounding boxes is identified from the plurality of bounding boxes after the second association. The set of unmatched bounding boxes are not associated with a tracker from the plurality of trackers during the second association. A third association is then performed, in which case each of the set of unmatched bounding boxes is associated with an associated tracker from the plurality of trackers that is within a second pre-determined distance.
-
公开(公告)号：US20180035120A1
公开(公告)日：2018-02-01
申请号：US15220856
申请日：2016-07-27
Applicant: QUALCOMM Incorporated
Abstract: In an example of this disclosure, a method may include receiving or generating image segmentation data for a plurality of pixels. The image segmentation data may include one or more image segmentation values for each pixel of the plurality of pixels. The method may include converting the one or more image segmentation values into one or more image component values. The method may include sending the one or more image component values corresponding to the one or more image segmentation values to a video encoder.
-
公开(公告)号：US12154189B2
公开(公告)日：2024-11-26
申请号：US17668956
申请日：2022-02-10
Applicant: QUALCOMM Incorporated
Inventor： Junkang Zhang , Zhen Wang , Lei Wang , Ning Bi
Abstract: Systems and techniques are described for image processing. An imaging system receives an identity image and an attribute image. The identity image depicts a first person having an identity. The attribute image depicts a second person having an attribute, such as a facial feature, an accessory worn by the second person, and/or an expression. The imaging system uses trained machine learning model(s) to generate a combined image based on the identity image and the attribute image. The combined image depicts a virtual person having both the identity of the first person and the attribute of the second person. The imaging system outputs the combined image, for instance by displaying the combined image or sending the combined image to a receiving device. In some examples, the imaging system updates the trained machine learning model(s) based on the combined image.
-
公开(公告)号：US11756334B2
公开(公告)日：2023-09-12
申请号：US17185811
申请日：2021-02-25
Applicant: QUALCOMM Incorporated
Inventor： Peng Liu , Lei Wang , Kuang-Man Huang , Michel Adib Sarkis , Ning Bi
IPC: G06V40/16 , G06N3/08 , G06F18/214
CPC classification number: G06V40/174 , G06F18/214 , G06N3/08 , G06V40/171
Abstract: Systems and techniques are provided for facial expression recognition. In some examples, a system receives an image frame corresponding to a face of a person. The system also determines, based on a three-dimensional model of the face, landmark feature information associated with landmark features of the face. The system then inputs, to at least one layer of a neural network trained for facial expression recognition, the image frame and the landmark feature information. The system further determines, using the neural network, a facial expression associated with the face.
-
-
-
-
-
-
-
-
-
Search Results
39
records
(took 0.038 seconds)
