In examples, a non-transitory computer-readable storage medium stores executable code, which, when executed by a processor, causes the processor to receive a semiconductor package image, the image including semiconductor package surface codes, the codes including a semiconductor package identifier. The executable code causes the processor to transmit at least one of the semiconductor package identifier, the codes, or the image. The executable code causes the processor to receive information associated with the semiconductor package identifier. The executable code causes the processor to output the information via at least one of a display coupled to the processor, a speaker coupled to the processor, or the wireless transceiver.

A method for maintaining inventory within a store includes: accessing an image (e.g., a color image, depth image) depicting an inventory structure; detecting a slot region of the image depicting a slot; identifying a product type assigned to the slot; accessing a product dimension of the product type; defining a target region within the slot in the image based on the product dimension; defining a product region within the slot in the image based on the product dimension and the target region; defining a back-of-shelf plane intersecting the target region of the image; detecting a surface within the product region; and, in response to the surface intersecting the back-of-shelf plane, identifying the slot as empty and generating a prompt to restock the slot with product units of the product type.

A method performed in a near sensor device (200) connected to a remote device (210) via a communication channel (220), for object detection in a video, the method comprising: detecting at least one object in the video scaled with a first set of scaling parameters (S312), using a first detection model (S314), encoding the video scaled with a second set of scaling parameters (S316), using an encoding quality parameter (S318), streaming the encoded video to the remote device (S320), streaming a side information associated to the encoded video to the remote device, wherein the side information comprises the information of the detected at least one object (S320), receiving a feedback from the remote device (S325), updating the configuration of the near sensor device (200) comprising adapting any of the first set of scaling parameters, the second set of scaling parameters, the first detection model and the encoding quality parameter (S340) based on the received feedback.

A system for continually regenerating adaptive, structured, reports in association with an image, the system comprising an imaging module (IM) to output images (I), a graphical user interface dashboard (GUI) to receive output images (I); a report regeneration module (RGT), an artificial imaging module (AIM) to graphically annotate each image (I) with a vector-defined boundary tag (VT) overlaid on the displayed image (I) and with to classify each image (I) with a classification label (CL), initializing a report (RP) to be generated and to be displayed, regenerating said initial report, to cause a first regenerated report, regenerating said first regenerated report, to cause an iteratively adapted regenerated report, to be displayed on said graphical user interface dashboard, said iteratively adapted regenerated report comprising pre-defined fields to be populated based on clinical diagnoses.

Multiple video files that are captured by calibrated imaging devices may be annotated based on a single annotation of an image frame of one of the video files. An operator may enter an annotation to an image frame via a user interface, and the annotation may be replicated from the image frame to other image frames that were captured at the same time and are included in other video files. Annotations may be updated by the operator and/or tracked in subsequent image frames. Predicted locations of the annotations in subsequent image frames within each of the video files may be determined, e.g., by a tracker, and a confidence level associated with any of the annotations may be calculated. Where the confidence level falls below a predetermined threshold, the operator may be prompted to delete or update the annotation, or the annotation may be deleted.

Storage (24) stores therein a teacher model (54) that is a target image (50) with a teacher annotation (52), which specifies the target image (50), assigned thereto. An inferring section (32) assigns based on the teacher model (54) an inferred annotation (82) to an inspection image (80) belonging to the same category to which the target image (50) belongs. The inferred annotation (82) is a result of inference from the inspection image (80) based on the teacher model (54). An aiding section (34) aids a human to determine whether or not the inferred annotation (82) specifies the inspection image (80). A correcting section (36) generates a corrected model (86) in a manner to add a corrected annotation (84) to the inferred annotation (82) so that the inspection image (80) is specified.

An image forming apparatus includes an image processing device, an image forming device, an operating device, a vital sensor, and a controller. The image processing device corrects an image. The image forming device performs an image formation of forming the image on a recording sheet. The operating device is operable by a user and through which an instruction to start the image formation by the image forming device is inputted. The vital sensor is provided at the operating device and detects a blood oxygen level of the user who is operating the operating device. The controller, when the instruction to start the image formation is inputted through the operating device, causes the image processing device to correct the image in accordance with the blood oxygen level of the user detected by the vital sensor and causes the image forming device to form a corrected image on the recording sheet.

A method and apparatus for correcting image data, an electronic device, a computer readable storage medium and an autonomous vehicle are provided. An embodiment of the method includes: acquiring auxiliary feedback data in response to absence of a target object in current feedback data, the target object being included in historical feedback data, collection time of the auxiliary feedback data being after collection time of the current feedback data, a difference between collection time of the historical feedback data and the collection time of the current feedback data being less than a first preset duration; extracting image data of the target object in response to the target object being included in the auxiliary feedback data; and correcting the current feedback data based on the image data.

A video annotation system for deep learning based video analytics and corresponding methods of use and operation are described that significantly improve the efficiency of video data frame labeling and the user experience. The video annotation system described herein may be deployed at a network edge and may support various intelligent annotation functionality including annotation tracking, adaptive video segmentation, and execution of predictive annotation algorithms. In addition, the video annotation system described herein supports team collaboration functionality in connection with large-scale labeling tasks.

Remotely surveying a photovoltaic system site includes receiving a photograph uploaded by a user device; analyzing the photograph using a trained machine learning model; receiving a confidence score from the trained machine learning model; determining if the photograph includes predetermined information, the predetermined information being used to perform a remote photovoltaic (PV) system site survey remotely; and in response to a determination that the photograph does not include the predetermined information, provide specific instructions regarding the missing information, wherein the specific instructions include guidance on how to retake the photograph to capture the predetermined information.