1st and 2nd FM generation 3D halftoning1 MICRON PIXELS133-MEGAPIXEL1-bit matrix completion180 degree images100 Hue test1/f noise108 Megapixel
2
2.5D reconstruction2D DCT2D VIEWS2D to Hologram conversion2-d scale2D metrics2D and 3D video2D-plus-depth video2.5 D printing2.5D PRINTING2D/3D imaging, high performance computing, imaging systems, efficient computations and storage2.5D printing2D AND 3D CONVERTIBLE DISPLAY2.5D2D2AFC2D-TO-3D CONVERSION ARTIFACTS2-D barcodes2D printing
3
3D Range Data3D surface reproduction3D STACK3D Modeling3D TRANSFORMATION3D compression3D Imaging3D model360-degree Image3D image compression3D Digital Image Correlation3D camera3D Halftoning3D Data Sources3D adaptive halftoning3D-HEVC360-degree videos3D connected tube model3D scene flow estimation3D Shape Indexing and Retrieval3D/4D DATA PROCESSING AND FILTERING3D Video Communications3D EDUCATIONAL MATERIAL3D Tracking3D Range Data Compression360 degree images3D modelling3D Mapping3D scanning3D capture3D Printing3D recovery3D affine transformation3D Point Cloud3D printing360-degree content3D Communications3D PRINTER3D shape3D HALFTONING360 IMAGING3D Curvelet3ARRI footage360° video3D range scanning3D depth sensing3D/2D Visuals3D modeling3D rigid transformations3D RECONSTRUCTION3D digital halftoning3D Vision3D Immersion3D Measurement3D theater program listing3D position measurement of people3D-shooting3D DIGITIZATION METHOD FOR OIL PAINTINGS3D Data Processing3D Telepresence3D shape indexing and retrieval3D mesh3D-color perception3D display3D refinement3D/4D Scanning3D video3D-LUT360-degree image projection3D CAMERAS360 degrees video360-degree art exhibition360 panorama3D perception3D stereo vision3D Morphable Model3D Object Detection360-degree images3D face alignment3D PRINTING3D MODEL360-video360-degree video streaming3D depth-map3D mapping and localization3D/4D Data Processing and Filtering360VR3D video processing3D scene classification3D objects3D warping3D Saliency3D surface3D data processing3D optical scans3DSR360-degree video3D communications3D displays3D MESHES3D Display360-degree3D Gaussian splatting3D VISUALIZATION3D Compression and Encryption3DMM3D reconstruction3D MESH360° STEREO PANORAMAS3D Iterative Halftoning3D Lidar3D Scene Reconstruction3D-CNN3D and 2D3D SCENE RECONSTRUCTION AND MODELING3D SHAPE INDEXING AND RETRIEVAL3D SALIENCY3D Meshes3-T pixel3D Video3D recursive search3D-human body detection3D halftoning3D print3D Models3D object shape3D Computer Graphics3D Compression360-DEGREE IMAGE360-deg quality assessment3D shape analysis3D Quality3D Video Conferencing3D RANGE IMAGING3D Image Processing3D MODELLING3-D RECONSTRUCTION3d localization3D/4D SCANNING3D localization3D-assisted features3DCNN3D-Anisotropic smoothing3D Scene Reconstruction and Modeling3D human-centered technologies3D point cloud3D mesh simplification35MM FILM DIGITIZATION3D projector3D ACQUISITION ARCHITECTURE3D audio3D imaging3D INTERACTION360° VIDEO3D printer3D RECOVERY3D colour Digital Image Correlation3D Print Appearance3D localization and mapping3D digitization and dissemination3D range geometry3D USER INTERFACES3D TV3D cinema and TV3D PROFILE360-Degree Video Technology3d mapping360x3D scene capture3-D SHAPE RECOVERY3A ALGORITHMS360 Video3D visual representation3D-high efficiency video coding3D DISPLAY3D STIMULI3d video3D IMAGE3D encoding3D3D vision3D glasses3D COMPRESSION AND ENCRYPTION3D surface structure based halftoning3D Reconstruction3D Color Printing360-degree imaging3D Range Data Encoding3d3D-printing
Although considerable progress has been made in recognizing multi-character text from images, there are still cases where there is a lack of robust computationally-efficient methods that can execute on portable devices to read device displays in the wild. We specifically address the problem of parsing digits from 7 segment displays. Recognizing these displays is important for many tasks such as assisting users with tasks using augmented reality agents that need to verify actions or connecting legacy devices to the internet for process control using cheap cameras. Legacy techniques based on image processing operators and OCR are brittle whereas massive deep networks are too computationally expensive. We describe a computationally tractable VGG style backbone combined with a novel digit inference head that can be trained using a synthetic display generator with novel augmentations. We show the model trained on augmented synthetic data generalizes well to a corpus of real-world display images getting 97.8% single-frame accuracy and obtaining a throughput of 30 frames per second. We describe how the output can be further stabilized to improve accuracy through a kind of mode filtering.