Qualcomm Incorporated

Young-Ki Baik

Duck Hoon Kim

Jeong-Kyun Lee

ChaeSeong Lim

Hyung-Il Koo

Hyun-Mook Cho

Hee-Seok Lee

Seungwoo Yoo

Seok-Soo Hong

Xinzhou Wu

Te-Won Lee

Taesu Kim

Kang Min Kim

Jayakrishnan Unnikrishnan

Ravi Teja Sukhavasi

Hankyu Cho

Kisun You

Minho Jin

Avdhut Joshi

Beom-su Kim

Minsub Lee

Yoga Y Nadaraajan

Amir Salimi

Gautam Sachdeva

Shivam Agarwal

Sree Sesha Aravind Vadrevu

Dongju Kim

Disclosed are techniques for estimating a 3D bounding box (3DBB) from a 2D bounding box (2DBB). Conventional techniques to estimate 3DBB from 2DBB rely upon classifying target vehicles within the 2DBB. When the target vehicle is misclassified, the projected bounding box from the estimated 3DBB is inaccurate. To address such issues, it is proposed to estimate the 3DBB without relying upon classifying the target vehicle.

Bounding box estimation and object detection

Bounding box estimation and lane vehicle association

Disclosed are techniques for determining a motion state of a target object. In an aspect, an on-board computer of an ego vehicle detects the target object in one or more images, determines one or more first attributes of the target object based on measurements of the one or more images, determines one or more second attributes of the target object based on measurements of a map of a roadway on which the target object is travelling, and determines the motion state of the target object based on the one or more first attributes and the one or more second attributes of the target object.

Determining a motion state of a target object

Disclosed are techniques for lane estimation. In aspects, a method includes receiving a plurality of camera frames captured by a camera sensor of a vehicle, assigning a first subset of the plurality of camera frames to a deep learning (DL) detector and a second subset of the plurality of camera frames to a computer vision (CV) detector based on availability of the DL and CV detectors, identifying a first set of lane boundary lines in a first camera frame processed by the DL detector, identifying a second set of lane boundary lines in a second camera frame processed by the CV detector, generating first and second sets of lane models based on the first and second sets of lane boundary lines, and updating a set of previously identified lane models based on the first set of lane models and/or the second set of lane models.

Hybrid lane estimation using both deep learning and computer vision

Techniques and systems are provided for determining features of one or more objects in one or more images. For example, an image of an object and a three-dimensional model associated with the object can be obtained. From the image, a sample point on the object can be determined. A depth and an angle of the sample point of the object can be determined. A pose and a shape of the three-dimensional model associated with the object can be determined based on the depth and the angle.

Point-based object localization from images

Disclosed are techniques for estimating multiple lane boundaries through simultaneous detection of lane markers (LM) and raised pavement markers (RPM). Conventional techniques for lane marker detection (LMD) comprises extracting and clustering line segments from a camera image, fitting the clustered lines to a geometric model, and selecting the multiple lanes using heuristic approaches. Unfortunately, in the conventional technique, an error from each step is sequentially propagated to the next step. Also there are no techniques to estimate both LMs and RPMs. A technique to simultaneously detect in real time both LM and RPMs is proposed. This enables optimal estimation of multiple lane boundaries.

Real-time simultaneous detection of lane marker and raised pavement marker for optimal estimation of multiple lane boundaries

A method, performed by an electronic device, for linking a thumbnail of an image and at least one web page is disclosed. In this method, the image including at least one text region may be accessed in a storage unit. At least one text region may be detected in the image and at least one character string in the at least one text region may be recognized. Further, the method may include selecting the at least one web page from the plurality of web pages and linking the thumbnail of the image and the at least one web page.

Linking thumbnail of image to web page

A method includes tracking an object in each of a plurality of frames of video data to generate a tracking result. The method also includes performing object processing of a subset of frames of the plurality of frames selected according to a multi-frame latency of an object detector or an object recognizer. The method includes combining the tracking result with an output of the object processing to produce a combined output.

Object tracking and processing

Techniques described herein provide a method for automatically and intelligently creating and updating an OCR cache while performing OCR using a computing device. An image captured using a camera coupled to the computing device may be matched against prior images stored in the OCR cache. If a match is found, the OCR cache may be updated with new or better information utilizing the new image. The matched prior image may be retained in the OCR cache, or the new captured image may replace the matched prior image in the OCR cache. In one embodiment, techniques are described to remove or reduce glare before storing the image in the OCR cache. In some embodiments, glare is removed or reduced in the absence of performing OCR.

Image cache

A method for communicating messages by a mobile device via a sound medium is disclosed. The mobile device receives input sounds from at least one mobile device via the sound medium. From the input sounds, an input sound signal carrying a first message encoded with a first key is detected. The mobile device decodes the first message based on a matching key. An output sound signal carrying a second message encoded with a second key is generated. Further, the mobile device transmits an output sound corresponding to the output sound signal via the sound medium.

Wireless communications using a sound signal

A portable computing device reads information embossed on a form factor utilizing a built-in digital camera and determines dissimilarity between each pair of embossed characters to confirm consistency. Techniques comprise capturing an image of a form factor having information embossed thereupon, and detecting embossed characters. The detecting utilizes a gradient image and one or more edge images with a mask corresponding to the regions for which specific information is expected to be found on the form factor. The embossed form factor may be a credit card, and the captured image may comprise an account number and an expiration date embossed upon the credit card. Detecting embossed characters may comprise detecting the account number and the expiration date of the credit card, and/or the detecting may utilize a gradient image and one or more edge images with a mask corresponding to the regions for the account number and expiration date.

Detecting embossed characters on form factor

The use of optical character recognition (OCR) in mobile devices is becoming prevalent with the increasing use of mobile devices. One important application for OCR in mobile devices is recognizing and translating the text to a language understandable by the user. Techniques are provided for replacing symbols in an image, while reducing the artifacts as a result of re-rendering of the background image.

Growing community of inventors

Seoul, South Korea

Young-Ki Baik