1-bit matrix completion180 degree images1/f noise108 Megapixel100 Hue test1st and 2nd FM generation 3D halftoning1 MICRON PIXELS133-MEGAPIXEL
2
2.5D PRINTING2D/3D imaging, high performance computing, imaging systems, efficient computations and storage2x2 On-chip Lens2.5 D printing2D AND 3D CONVERTIBLE DISPLAY2.5D printing2AFC2.5D2D2D printing2D-TO-3D CONVERSION ARTIFACTS2-D barcodes2D DCT2D VIEWS2.5D reconstruction2-d scale2D metrics2D to Hologram conversion2D-plus-depth video2D and 3D video
3
3D MESH3D Iterative Halftoning360° STEREO PANORAMAS3D Scene Reconstruction3D Lidar3D-CNN3DMM3D VISUALIZATION3D Compression and Encryption3D reconstruction3D Display3D MESHES360-degree3D Gaussian splatting3DSR3DViewers3D communications360-degree video3D displays360-deg quality assessment360-DEGREE IMAGE3D shape analysis3D Quality3D Video Conferencing3D Image Processing3D RANGE IMAGING3-D RECONSTRUCTION3D MODELLING3d localization3D Models3D Computer Graphics3D Compression3D object shape3D Meshes3D SALIENCY3D Video3-T pixel3D recursive search3D-human body detection3D halftoning3D print3D and 2D3D SCENE RECONSTRUCTION AND MODELING3D SHAPE INDEXING AND RETRIEVAL3D Print Appearance3D colour Digital Image Correlation3D digitization and dissemination3D localization and mapping3D range geometry3D imaging3D audio3D ACQUISITION ARCHITECTURE360° VIDEO3D INTERACTION3D printer3D Photogrammetry3D RECOVERY3D-Anisotropic smoothing3DCNN3D Scene Reconstruction and Modeling3D human-centered technologies3D point cloud35MM FILM DIGITIZATION3D mesh simplification3D projector3D/4D SCANNING3D localization3D-assisted features3D Reconstruction3D Color Printing360-degree imaging3D CG Image3d3D Range Data Encoding3D-printing3D STIMULI3A Algorithms3d video3D IMAGE3D encoding3D3DGS3D vision3D surface structure based halftoning3D COMPRESSION AND ENCRYPTION3D glasses3d mapping3-D SHAPE RECOVERY360x3D scene capture3A ALGORITHMS360 Video3D visual representation3D-high efficiency video coding3D DISPLAY3D USER INTERFACES3D TV3D cinema and TV360-Degree Video Technology3D PROFILE3D connected tube model3D scene flow estimation360-degree videos3D/4D DATA PROCESSING AND FILTERING3D Shape Indexing and Retrieval3D Video Communications3D EDUCATIONAL MATERIAL3D Tracking3D image compression3D Digital Image Correlation3D camera3D Data Sources3D Halftoning3D adaptive halftoning3D-HEVC3D TRANSFORMATION3D Imaging3D compression3D model360-degree Image3D Range Data3D surface reproduction3D STACK3D Modeling3D/2D Visuals3D skeletal joint point cloud; deep learning; PointNet++ method3D modeling3D RECONSTRUCTION3D rigid transformations3D Immersion3D Vision3D digital halftoning3D HALFTONING360 IMAGING360° video3ARRI footage3D Curvelet3D range scanning3D depth sensing3D printing360-degree content3D Communications3D PRINTER3D shape3D Range Data Compression360 degree images3D modelling3D capture3D scanning3D Mapping3D Printing3D recovery3D affine transformation3D Point Cloud3D Morphable Model360-degree images3D Object Detection3D MODEL3D face alignment3D PRINTING3D/4D Scanning3D-LUT3D video360-degree image projection360 degrees video360-degree art exhibition3D CAMERAS3D perception360 panorama3D stereo vision3D shape indexing and retrieval3D mesh3D display3D-color perception3D refinement3D theater program listing3D Measurement3D-shooting3D position measurement of people3D DIGITIZATION METHOD FOR OIL PAINTINGS3DLUT3D Telepresence3D image3D Data Processing3D surface3D Saliency3D Gaussian Splatting3D optical scans3D data processing3D video processing3D scene classification3D objects3D warping3D mapping and localization360VR3D/4D Data Processing and Filtering360-video3D depth-map360-degree video streaming
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.