A medical device monitoring system and method extract information from screen images from medical device controllers, with a single OCR process invocation per screen image, despite critical information appearing in different screen locations, depending on which medical device controller's screen image is processed. For example, different software versions of the medical device controllers might display the same type of information in different screen locations. Copies of the critical screen information, one copy from each different screen location, are made in a mosaic image, and then the mosaic image is OCR processed to produce text results. Text is selectively extracted from the OCR text results, depending on contents of a selector field on the screen image, such as a software version number or a heart pump model identifier.

Computer vision systems and methods for text classification are provided. The system detects a plurality of text regions in an image and generates a bounding box for each detected text region. The system utilizes a neural network to recognize text present within each bounding box and classifies the recognized text, based on at least one extracted feature of each bounding box and the recognized text present within each bounding box, according to a plurality of predefined tags. The system can associate a key with a value and return a key-value pair for each predefined tag.

A customer premises parking system including one or more cameras and wireless communications transmission capability is described. The presence of a vehicle with a license plate at an entrance or exit of a parking area is detected and one or more high resolution images are captured and time stamped. A license plate portion of the HD image is captured and stored. The license plate number and corresponding time is determined and communicated using a wide area wireless network. The HD license plate portion of the image is processed to produce a low resolution version of the license plate image which is segmented and processed into fragments which are transmitted via the wide area network using available transmission opportunities. A high resolution version of the license plate portion is stored at the premises where the parking facilitate is located and can be retrieved via a cellular wireless interface in the event of a dispute or discrepancy with regard to the detected and reported license plate number. From the time stamps associated with entrance and exit images of license plates the amount of time and/or date at which a vehicle was present at the parking garage is determined and the vehicle owner billed for the use of the parking facility.

A customer premises parking system including one or more cameras and wireless communications transmission capability is described. The presence of a vehicle with a license plate at an entrance or exit of a parking area is detected and one or more high resolution images are captured and time stamped. A license plate portion of the HD image is captured and stored. The license plate number and corresponding time is determined and communicated using a wide area wireless network. The HD license plate portion of the image is processed to produce a low resolution version of the license plate image which is segmented and processed into fragments which are transmitted via the wide area network using available transmission opportunities. A high resolution version of the license plate portion is stored at the premises where the parking facilitate is located and can be retrieved via a cellular wireless interface in the event of a dispute or discrepancy with regard to the detected and reported license plate number. From the time stamps associated with entrance and exit images of license plates the amount of time and/or date at which a vehicle was present at the parking garage is determined and the vehicle owner billed for the use of the parking facility.

Some embodiments provide a method that predicts data items from a real-world object in real-time. The method captures a video comprising a plurality of frames. The method further performs object detection on a frame in the plurality of frames to determine that the frame includes an object. The method also processes the frame using a plurality of models, wherein each model in the plurality of models is configured to predict a set of candidate data items associated with the object. The method selects one or more candidate data items from the sets of candidate data items associated with the object as a set of data items. The method populates a record with the set of data items.

Some embodiments provide a method that predicts data items from a real-world object in real-time. The method captures a video comprising a plurality of frames. The method further performs object detection on a frame in the plurality of frames to determine that the frame includes an object. The method also processes the frame using a plurality of models, wherein each model in the plurality of models is configured to predict a set of candidate data items associated with the object. The method selects one or more candidate data items from the sets of candidate data items associated with the object as a set of data items. The method populates a record with the set of data items.

A system and a method for training an inference model using a computing device. The method includes: providing a text-to-vector converter; providing the inference model and pre-training the inference model using labeled fashion entries; providing non-labeled fashion entries; separating each of the non-labeled fashion entries into a target image and target text; converting the target text into a category vector and an attribute vector using the text-to-vector converter; processing the target image using the inference model to obtain processed target image and target image label; comparing the category vector to the target image label; when the category vector matches the target image label, updating the target image label based on the category vector and the attribute vector to obtain updated label; and retraining the inference model using the processed target image and the updated label.

A system and a method for training an inference model using a computing device. The method includes: providing a text-to-vector converter; providing the inference model and pre-training the inference model using labeled fashion entries; providing non-labeled fashion entries; separating each of the non-labeled fashion entries into a target image and target text; converting the target text into a category vector and an attribute vector using the text-to-vector converter; processing the target image using the inference model to obtain processed target image and target image label; comparing the category vector to the target image label; when the category vector matches the target image label, updating the target image label based on the category vector and the attribute vector to obtain updated label; and retraining the inference model using the processed target image and the updated label.

A method of operating a carwash includes driving a vehicle to a point of sale station of the carwash, purchasing a carwash transaction, and taking a first picture of a license plate number of the vehicle after the completion of the purchase of the carwash transaction at the point of sale station. The license plate number is associated with the carwash transaction as a carwash ticket. The vehicle is then sent to a carwash queue disposed downstream of the point of sale station. The carwash ticket information is sent to a carwash tunnel. A second picture of the license plate number of the vehicle is taken right as it approaches the carwash tunnel. The license plate number is associated with the carwash ticket and the vehicle is washed according to a wash dictated by the carwash ticket.

A method of operating a carwash includes driving a vehicle to a point of sale station of the carwash, purchasing a carwash transaction, and taking a first picture of a license plate number of the vehicle after the completion of the purchase of the carwash transaction at the point of sale station. The license plate number is associated with the carwash transaction as a carwash ticket. The vehicle is then sent to a carwash queue disposed downstream of the point of sale station. The carwash ticket information is sent to a carwash tunnel. A second picture of the license plate number of the vehicle is taken right as it approaches the carwash tunnel. The license plate number is associated with the carwash ticket and the vehicle is washed according to a wash dictated by the carwash ticket.