Abstract: Computer vision grouping recognition is provided by receiving training images that include unpackaged items; identifying, by a computer vision model, a candidate identities for unpackaged items in a given training image; receiving, from a human user, a selected identity for the unpackaged item as feedback for the candidate identity; constructing a confusion matrix tallying matches and mismatches between candidate identities and the selected identities as analyzed across the training images for each unpackaged item; identifying at least one product category that includes at least a first unpackaged item and a second unpackaged item that the confusion matrix indicates as being misidentified for each other by the computer vision model; and reconfiguring the computer vision model to identify the product category instead of the first unpackaged item or the second unpackaged item when analyzing a given image including one or more of the first unpackaged item or the second unpackaged item.

Abstract: Computer vision grouping recognition is provided by receiving training images that include unpackaged items; identifying, by a computer vision model, a candidate identities for unpackaged items in a given training image; receiving, from a human user, a selected identity for the unpackaged item as feedback for the candidate identity; constructing a confusion matrix tallying matches and mismatches between candidate identities and the selected identities as analyzed across the training images for each unpackaged item; identifying at least one product category that includes at least a first unpackaged item and a second unpackaged item that the confusion matrix indicates as being misidentified for each other by the computer vision model; and reconfiguring the computer vision model to identify the product category instead of the first unpackaged item or the second unpackaged item when analyzing a given image including one or more of the first unpackaged item or the second unpackaged item.

Abstract: Computer vision grouping recognition is provided by receiving training images that include unpackaged items; identifying, by a computer vision model, a candidate identities for unpackaged items in a given training image; receiving, from a human user, a selected identity for the unpackaged item as feedback for the candidate identity; constructing a confusion matrix tallying matches and mismatches between candidate identities and the selected identities as analyzed across the training images for each unpackaged item; identifying at least one product category that includes at least a first unpackaged item and a second unpackaged item that the confusion matrix indicates as being misidentified for each other by the computer vision model; and reconfiguring the computer vision model to identify the product category instead of the first unpackaged item or the second unpackaged item when analyzing a given image including one or more of the first unpackaged item or the second unpackaged item.

Abstract: Because a touch controller may not be capable of providing a signal to indicate when to provide audio feedback, a computing system may include a touch event detector that is located in a communication channel between a touch controller and a processing system (e.g., a processor and main memory of a computing system). The touch event detector may monitor the touch data as it is relayed from the touch controller to the processing system via a communication hub. Once the touch event detector identifies a relevant touch event (i.e., a touch event that should provide audio feedback to a user), the touch event detector transmits an instruction to an audio output device (e.g., a speaker) which then generates the audio feedback.

Abstract: Multiple users may interact with the multiple touch monitors simultaneously. For example, a point-of-sale (POS) system may include two touch monitors (e.g., one facing the cashier and another facing the customer) that are controlled by the same operating system. To provide audio feedback to both users, the operating system needs to know on which monitor the touch event occurred. However, because the operating system may use a shared coordinate region, it only knows that a touch event occurred within the region but not know on which monitor. As such, embodiments herein describe a filter module that identifies a specific location of a touch event within the shared coordinate region and maps that location to one of the touch monitors. Once the specific touch monitor is identified, the filter module sends an instruction that causes a speaker assigned to the identified touch monitor to output the audio feedback.

Abstract: A circuit is attached in parallel to a universal serial bus interface of a data processing system. A capacitor in the circuit is charged by receiving power from a power pin of the universal serial bus interface while the data processing system is not in a reduced power state. A vibration sensor is unpowered while the data processing system is not in a reduced power state. The vibration sensor is disconnected from a data pin of the universal serial bus interface while the data processing system is not in a reduced power state. When the data processing system enters a reduced power state, the capacitor provides power to the vibration sensor. When a vibration is detected by the vibration sensor, a switch connects the vibration sensor to the data pin of the universal serial bus interface, providing a wake up signal to the data processing system.

Abstract: A circuit is attached in parallel to a universal serial bus interface of a data processing system. A capacitor in the circuit is charged by receiving power from a power pin of the universal serial bus interface while the data processing system is not in a reduced power state. A vibration sensor is unpowered while the data processing system is not in a reduced power state. The vibration sensor is disconnected from a data pin of the universal serial bus interface while the data processing system is not in a reduced power state. When the data processing system enters a reduced power state, the capacitor provides power to the vibration sensor. When a vibration is detected by the vibration sensor, a switch connects the vibration sensor to the data pin of the universal serial bus interface, providing a wake up signal to the data processing system.

Abstract: A touch screen device comprising a display screen, an electromagnetic digitizer disposed behind the display screen for detecting stylus proximity and stylus coordinates, and an infrared touch device secured in front of the display screen for detecting the coordinates of an infrared opaque object. The touch screen device also includes a controller in electronic communication with the electromagnetic digitizer and the infrared touch device. The controller provides an output signal based upon the infrared opaque object coordinates detected by the infrared touch device during periods of time that the electromagnetic digitizer is not detecting stylus proximity and provides an output signal based upon the stylus coordinates detected by the electromagnetic digitizer during periods of time that the electromagnetic digitizer is detecting stylus proximity.

Abstract: A keyboard is placed into a suspend mode. Upon the detection of the pressing of a key, the keyboard processor is wakened from the suspend mode. No power is consumed during the suspend mode, sine the keyboard processor is not scanning the key matrix. The key matrix drive circuitry may also be tested by monitoring a signal emanating from the key matrix drive circuitry and scanning each of the drive lines in the key matrix. If the signal is altered, then the associated drive line in the keyboard drive circuitry is defective. Testing of sense circuitry is performed by changing pull-up resistors to pull-down resistors and then reading sense lines.