A version of nrsc5-gui that merges the interface developed by cmnybo with the architecture developed by zefie in order to start a new baseline that is not heavily dependent upon Python processing.

Overview

NRSC5-DUI is a graphical interface for nrsc5. It makes it easy to play your favorite FM HD radio stations using an RTL-SDR dongle. It will also display weather radar and traffic maps found on most iHeart radio stations.

This version is really a fork of a fork of the original nrsc5-gui: The first was developed by cmnybo and subsequently modified by zefie. It merges the features of the former to the architecture of the latter, while adding several additional control and display features.

As such, we have changed the name to 'DUI' as a play on the Italian word for 'two', this being a second generation graphical user interface for nrsc5. (I'll be here all week. Please tip your waitresses.)

Dependencies

The following programs are required to run NRSC5-DUI

Setup

  1. Install the latest version of Python 3.9, PyGObject, Pillow, and other dependencies.
  2. Compile and install nrsc5.
  3. Install sox
  4. Install nrsc5-dui files in a directory where you have write permissions.

The configuration and resource directories will be created in a new cfg and res directory under where nrsc5-dui.py resides. Similarly, an aas directory will be created for downloaded files and a map directory will be created to store weather & traffic maps.

nrsc5 should be installed in a directory that is in your $PATH environment variable. Otherwise add the full path to nrsc5 (e.g., /usr/local/bin/) may be entered at runtime.

Usage

Please ensure your RTL-SDR dongle is first connected to an available USB port. Then, from the terminal, start nrsc5-dui by entering: python3 nrsc5-dui.py or something like: python3 nrsc5-dui.py /usr/local/bin/ to include the path to nrsc5 when using scripts (like Apple Script) that seemingly ignore the environment.

Settings

You may first change some optional parameters of how nrsc5 works from the Settings tab in nrsc5-dui: Set the gain to Auto, or optionally enter an RF gain in dB that has known to work well for some stations.
Enter a PPM correction value if your RTL-SDR dongle has an offset.
Enter the IP address that rtl_tcp is listening to and check the Enabled box if you are using a remote RTL-SDR.
Enter the number of the desired device if you have more than one RTL-SDR dongle.
Check Log to file to enable writing debug information from nrsc5 to nrsc5.log.
Check Download Album Art to enable automated downloading of album art from MusicBrainz.
Check Include Station Art to display album art that is generated by the station in addition to downloading from MusicBrainz.
Check Extended Queries to apply several MusicBrainz queries to find album art. Turning this option on may be slower than non-extended queries.

Playing

Enter the frequency in MHz of the station you want to play and either click the triangular Play button on the toolbar, or just hit return. When the receiver attains synchronization, the pilot in the lower left corner of the status bar will turn green. It will return to gray if synchronization is lost. If the device itself becomes 'lost', the pilot will turn red to indicate an error has occurred (this is the theory, though I've yet to see this status message happen in practice). The synchronization process may take about 10 seconds, and the station will begin to play. This depends upon signal strength and whether it's relatively free from adjacent interference. After a short while, the station name will appear to the right of the frequency, and the available streams will show on a row of buttons just beneath the frequency entry. Clicking one of these buttons will change to that particular stream. Note: No settings other than stream may be changed while the device is playing.

Album Art & Track Info

Some stations will send album art and station logos. These will fill the Album Art tab, as they are made available by the station. Most stations will send the song title, artist, album, and genre. These are displayed in the Track Info pane, also if available. The user can override what the stations send by enabling the DL Album Art setting. This will use the Title and Artist information to retrieve album art from MusicBrainz. If no album art is found, the station logo will be used, if available. The title, artist, album, and genre (if available) will be cached when new album art is found, and will be automatically displayed when that art is used. The user can change the logo of the playing station by right-clicking in the Album Art area. This will display a popup prompting you for the URL of an image found on the web. Pasting the URL in the box and clicking 'Ok' will download the image and set the logo of the playing station with it.

Bookmarks

When a station is playing, you can click the Bookmark Station button to add it to the bookmarks list. You can click on the Name in the bookmarks list to edit it. Double click the Station to tune to that particular station and stream. Click the Delete Bookmark button to delete it. Note that some stations use the default MPS/SPS or HDn naming for their streams. In this case, the bookmark will be used to name the stream button.

Station Info

The station name, slogan, message, and optional alert message will display if the station as pre-programmed them. The current audio bit rate will be displayed here as well as on the status bar. The stations available streams and data services, with a description of each will display, as the station has pre-programmed them. This is a useful feature for noting which stations have Total Traffic & Weather Network traffic and weather images.

Signal Strength

The Modulation Error Ratio for the lower and upper sidebands are displayed as they are determined. Important: High MER values for both sidebands indicates a strong signal. The current, average, minimum and maximum Bit Error Rates will also be displayed as they are determined. High BER values will cause the audio to glitch or drop out. The current BER is also shown on the status bar and may be used as a tuning tool.

Maps

When listening to radio stations operated by iHeartMedia, you may view live traffic maps and weather radar. The images are typically sent every few minutes and will fill the tab area once received, processed, and loaded. Clicking the Map Viewer button on the toolbar will open a larger window to view the maps at full size. The weather radar information from the last 12 hours will be stored and can be played back by selecting the Animate Radar option. The delay between frames (in seconds) can be adjusted by changing the Animation Speed value. Other stations provide Navteq/HERE navigation information... it's on the TODO 'like to have' list.

Map Customization

The default map used for the weather radar comes from OpenStreetMap. You can replace the map.png image with a map from any website that will let you export map tiles. The tiles used are (35,84) to (81,110) at zoom level 8. The image is 12032x6912 pixels. The portion of the map used for your area is cached in the map directory. If you change the map image, you will have to delete the BaseMap images in the map directory so they will be recreated with the new map.

Screenshots

album art tab info tab settings tab bookmarks tab map tab

Version History

1.0.0 Initial Release
1.0.1 Fixed compatibility with display scaling
1.1.0 Added weather radar and traffic map viewer
1.2.0 zefie update to modern nrsc5 build
2.0.0 Updated to use the nrsc5 API
2.1.0 Updated and enhanced operation and use

Comments
  • Testing Notes

    Testing Notes

    Thank you for doing this! The audio works very well.

    My System- Ubuntu 20.04

    I do have only one issue I see right now and it's this

    Mar 27 17:22:02 : Error: Unable to load station logo database
    Mar 27 17:22:02 : Error: Unable to load config
    Mar 27 17:22:43 : Error: Unable to save config
    

    Not sure if this is a WIP still but I cannot save favorites or save the station logos. The album art works fine though. Do you know any fixes? besides that it works awesome!

    opened by andrewfer000 82
  • Support for NRSC5 executables built for SDRplay functionality?

    Support for NRSC5 executables built for SDRplay functionality?

    Hi, I use an SDRplay RSPdx device instead of the RTL ones and was able to build and use a forked version of nrsc5 that supports it no problem. A few of the command-line arguments are a little different than the RTL ones though and I wondered if you might be able to add a few tweaks so that nrsc5-dui can spit out the right flags for device id and allow you to tweak the gain settings? I don't know if you are still actively working on the project, but just figured i'd ask! nrsc5-dui is an awesome program!

    opened by dragonbytes 55
  • Windows Support Thread

    Windows Support Thread

    Creating this as a thread for RTL-SDR Users on Microsoft Windows Operating Systems. Since neither myself or Mark use Windows, we will be relying on other users to test and give issues. Based on what I know, you just need to install python 3.8 or 3.9 and install the required modules from Pip, then install nrsc5 for Windows and try and run the application. You might have to change the /usr/local/bin/nrsc5 hard-coded path to the nrsc5.exe file installed on Windows or the one in the nrsc5-dui folder. We could really use some Windows testers, so if you are using that OS and are skilled enough with how python works then feel free to help out!

    opened by andrewfer000 51
  • Simple Fix for Windows users/WSL2

    Simple Fix for Windows users/WSL2

    I found a very simple fix to get your app running under Windows/WSL2 without any issues. No need for pulseaudio or complicated USB pass-through solutions. WSL2 accesses the Windows file system from the /mnt folder. I.e., to access files on the Windows C: drive, provide the following path: /mnt/c and to access the D: drive, /mnt/d/ and so on.

    By editing line 74 of nrsc5-dui.py and adding the path for the nrsc5.exe and libnrsc5.dll, the app runs fine. The edited line reads: self.nrsc5Path = arg1+"/mnt/d/nrsc5/nrsc5.exe"

    My EXE and DLL are in my D: drive in a folder named nrsc5. For users who have their nrsc5 files in another location, simply edit line 74 to reflect where your files are stored.

    I have to invoke the app as root using sudo, like so: sudo python3 ./nrsc5-dui.py Not sure why, but without sudo, python will throw an error.

    Here are some screenshots:

    image

    image

    image

    opened by ferrellsl 16
  • Need help with MacOS

    Need help with MacOS

    Looking for help! Running MacOS Monterey Beta 10. nrsc5 compiles and runs like a champ. No issues here.

    nrsc5-dui will not run. I carefully followed the steps given in the Read Me install section. Get this message and a Black Screen with a Title Bar upon launch. Here are the first two error messages. I'm lost!

    (nrsc5-dui.py:75486): Gtk-WARNING **: 01:57:37.670: Could not load a pixbuf from /org/gtk/libgtk/theme/Adwaita/assets/check-symbolic.svg. This may indicate that pixbuf loaders or the mime database could not be found.

    (nrsc5-dui.py:75486): Gtk-WARNING **: 01:57:37.690: Error loading theme icon 'media-playback-start' for stock:

    Any help would be appreciated.

    73 de Dale - K0HYD

    help wanted 
    opened by K0HYD 13
  • Getting more unusable on small screens

    Getting more unusable on small screens

    Hey Mark,

    As I mentioned in a past issue at Nrsc5 DUI is getting harder to use on Lower resolution displays such as on my old laptop.

    I think there's a redesign or second layout option needed that supports wider window with shorter window height instead of this vertical style alone.

    Obviously this is an edge case so it's up to you if you think it's worth implementing. Maybe having a short and wide layout that can be enabled with a command line switch would be a good option.

    enhancement 
    opened by andrewfer000 11
  • (Question) Does this support rtl_tcp

    (Question) Does this support rtl_tcp

    I run a personal rtl_tcp server to get my antenna in a better reception spot in my house, while still using RTL-SDR anywhere around the house, and i would like to know if this supports it.

    opened by grumpzalt 10
  • Installing nrsc5-dui on Windows 11

    Installing nrsc5-dui on Windows 11

    Hi, I understand that at this point of time, there isn't an official .exe version of nrsc5-dui for Windows computers, so yesterday I tried to compile the source code for nrsc5-dui in MSYS2 MinGWx64 and MSYS2 MinGWx86. I keep running into this problem where the setup.py file can't find a module called gtk+ and I know that I have gtk2 and gtk3 installed in MSYS2 MinGWx64, so I'm not sure how to fix this error message. image

    opened by Fairyball 5
  • Unable to load audio device

    Unable to load audio device

    I get the error 21:16:15 Unable to open audio device. when using it

    Also isn't able to connect to my RTL, whilst the regular nrsc5-gui doesn't have any issues connecting.

    Using Kubuntu 20.04

    opened by CallumCat 4
  • Feature Request/Dark Mode

    Feature Request/Dark Mode

    Would it be possible to implement a dark mode? The white window background can be particularly eye-searing in dim settings. Maybe a black background with light gray text. Thanks for all the work you've put into this app. I've enjoyed it immensely. I set aside an entire mini-HTPC as my HD Radio infotainment center and placed it in my kitchen.

    IMG_20220217_104836

    opened by ferrellsl 3
  • Deprecation Errors

    Deprecation Errors

    image

    Something I'd like to see fixed or suppressed is the constant console spam from the Gtk libraries.

    I usually leave the console open in the background in order to see and determine if there's any errors, but when this is constantly spamming me it would be nice to fix it.

    opened by CallumCat 3
  • Feature Request/Docker container?

    Feature Request/Docker container?

    I love this app. Got it working great on an Ubuntu VM on my Windows 11 box using VMWare. Since running this app natively on Windows is problematic at best, would it be possible to create a Docker app to preclude running this on a VM?

    enhancement 
    opened by ferrellsl 1
  • Possible to run without an audio device?

    Possible to run without an audio device?

    I'd like to capture to files, and the dusty machine nearest to where I can plug in the antenna doesn't have an audio device. Is there a way to configure the app to only dump to audio files?

    I did get it to compile and run in Ubuntu, and got as far as

    $ more nrsc5.log 
    15:32:17 Unable to open audio device.
    
    enhancement 
    opened by salamanders 2
  • Analog (analogue) Radio Implmentation [Long Term Goal]

    Analog (analogue) Radio Implmentation [Long Term Goal]

    Starting this issue in case someone starts working or wants to start working on implementing the following features and want's a good place to document problems and progress. This is somewhat extended from the original goal but not by much.

    • [ ] Broadcast FM Support with RDS
    • [ ] Broadcast AM Support with C-QUAM
    • [ ] SW Radio support (would just require a change of frequency range of AM)
    • [ ] DRM Support for AM/SW (I'm assuming)
    • [ ] NOAA Weather Radio (Narrow Band FM, 162-164Mhz)
    • [ ] Scanning functionality

    I think building it off of something like rtl_fm might be a good starting point. Since Nrsc5-DUI is just a frontend, it's better to have a way to do the demodulation and audio processing outside of the python program.

    enhancement 
    opened by andrewfer000 8
  • TODO List

    TODO List

    • [ ] 1. Code Cleanup. I've been fooling around with the app on Linux as well and seems like it's doing what it's supposed to do. I'd really like to start cleaning up the code style and make it look closer to 2.0... given time.
    • [x] 2.a. Stream Names: I'm a little disappointed in some of these stations' sloppy implementation. Naming a stream MPS/SPS1/SPS2 is pretty lazy. Think I'll write something that looks in the bookmarks and uses the name in there as an alternative.
    • [x] 2.b. Stream buttons or labels should not change the stream if they have no text. No point in selecting a stream that is not used.
    • [x] 3.a. Logos: Right now, the stations broadcast logos for each stream, but the app only applies it to the logo list if you're tuned to it when it's received. I understand the reasoning for that, but think it should be applied when received whether you're turned to the stream or not.
    • [x] 3.b. Investigate why new logos aren't showing unless already cached in stationLogos.json, requiring a restart.
    • [x] 3.c. Added quick capability to change the station logo of the playing station.
    • [ ] 4.a. Covers: Modify the way the album art works as a repository to move and consolidate the images out of the main aas directory that can be recalled later when the ID3 info changes. Hierarchy should be set up as: Artist, album, and song (the latter solely for image reference purposes).
    • [x] 4.b. Also provide a means for the user to correct an image using an online database such as Discogs.
    • [x] 4.c. Investigate why/how Discogs is now 403-ing automated searches and/or find another source - Now coming from MusicBrainz via the musicbrainzngs pip3 module.
    • [x] 4.d. Need to fix downloading of cover images using musicbrainzngs.
    • [x] 4.e. Add ability to display or not display any station-generated cover art when downloading from MusicBrainz.
    • [ ] 4.f. Add facility to manually import cover art in case automated download fails: Add a magifying glass button on the toolbar. When clicked, it displays a dialog with the Title and Artist. Button on dialog spawns a web browser with the appropriate MusicBrainz query. Copy/Paste the URL for the image you want, as well as the applicable Album and Genre metadata, and click Ok (or Cancel). Image would then be downloaded to the repository and metadata is cached in the database as if it was an automated download.
    • [ ] 5. Album Art Tab Redesign: Also thinking about redesigning the Album Art tab to look more like how a car radio might actually look. Maybe split the area into a line at the top that shows the station call and small image of the logo in the upper right (or to the right of where I have the call showing now), ~~a couple of small lines at the bottom that show the slogan and message~~, and then flash the covers as they come in the center between them. Some don't have covers, so that would be a larger static version of the logo. ~~If I find a way to do it, I'll put the slogan and message on one line, and make it crawl across the bottom.~~
    • [ ] 6. Alternate Nav Data: I want to research how Navteq/HERE works and see what can be done with it. Seems like another source of info in case an iHeart station isn't available. There are other digital services that are used that can also be taken advantage of.
    • [x] 7. Traffic Timestamps: I have no idea why they left the timestamp off the traffic maps. That thing could be five days old and you'd never know it. At the least I'll add the timestamp, or have something on there that displays the dtg that the file was saved so it doesn't obscure part of the map.
    • [ ] 8. Use With Other Radios: I'd like to modify nrsc5 to work with my AirSpy HF Discovery, maybe my first priority. The RTL-SDR is a nice toy, and it does a lot of things, but there are a lot of problems trying to use it for something like this. There are a lot of other stations that I'm sure an AirSpy would be able to pick up that are otherwise getting washed out on an RTL-SDR. I put a feeler out to see if anything's being done to use other receivers, but that kind of fell flat.
    • [ ] 8.a. AirSpy HF+ Discovery
    • [x] 8.b. SDRPlay (various models)
    • [ ] 9. Long-term goal: Expand operation past nrsc5 into other modes for broadcast and utility listening.
    • [ ] 9.a. Add Other Modes: Add standard FM Stereo with RDS, AM with C-QUAM, SW with DRM.
    • [ ] 9.b. Add VHF utility support: Weather band radio, scanning, decoding of trunked transmissions, etc.
    • [x] 10. NRSC5 Path: Automate finding the absolute path to nrsc5. Hardcoding line 68 of nrsc5-dui.py won't work for everyone.
    • [x] 11. Redundant Files: Remove redundant files in the repository to reduce clutter and confusion. No need to have images and glade files in root, when the app looks for them in res
    • [x] 12. Ellipcize and add tooltips to ID3 info (title, artist, album, genre) in case text exceeds display limitations.
    • [x] 13. Eliminate threading and stock icon deprication warnings.
    • [x] 14. Detect if RTL-SDR is actually connected before starting nrsc5.
    • [x] 15. Add feature to run nrsc5 remote via rtl_tcp (-H <IP>). Save setting and IP to cfg for next run.
    • [ ] 16. Convert for Win11 operation.
    • [x] 17. Add an alternate layout with eight buttons in one row at the top, with the body underneath consisting of the tabbed section on the left and the id3 details on the right.
    • [ ] 18. Add a Docker container for Linux installations.
    • [ ] 19. Add a capability to read/write IQ files as well as to save audio to wav files in lieu of a PC sound system.
    enhancement 
    opened by markjfine 0
Releases(v2.2.2)
  • v2.2.2(Dec 19, 2022)

    Added Auto antenna setting for SDRPlay. Removed quotes from SDRPlay antenna command line setting. Fixed an issue regarding how the gain setting was being implemented on SDRPlay.

    Source code(tar.gz)
    Source code(zip)
  • v2.2.1(Dec 15, 2022)

  • v2.2.0(Dec 9, 2022)

    This release adds the capability to use SDRPlay as well as the full complement of 8 nrsc5 streams per station. It also cleans up a few interface nits and stability issues.

    Full Changelog: https://github.com/markjfine/nrsc5-dui/compare/v2.1.1...v2.2.0

    Source code(tar.gz)
    Source code(zip)
  • v2.1.1(May 11, 2021)

  • v2.1.0(May 7, 2021)

    First official release of nrsc5-dui, although calling it v2.1.0 to maintain continuity with it's predecessor, nrsc5-gui.

    This release adds several new desirable features, such as:

    1. User-responsive controls: The tuning interface has been redesigned to allow the user to type in a frequency and hit return to tune, as well as select one of four titled stream buttons.
    2. Display of additional station-provided data and status: The overall display has added Album and Genre for when such track data is provided by the station (or MusicBrainz, see 3, below). The bottom status area has also been modified to include a receiver status pilot and changes the bit error rate from 'average' to 'current' in order to be more responsive for antenna tuning. The Station Info tab has been expanded to include the gamut of what a station could possibly provide (for those who are nosy, like me), including audio and data stream names and types. The latter is likewise helpful to determine if the station provides TTN traffic and weather imagery. The audio labels are tunable, in addition to the buttons up top, so clicking on any of them will change the playing stream.
    3. Display of additional station-related imagery: The user may add station logos when the station doesn't provide one. It will also use the track information to automatically download and reuse album art and additional Album and Genre information from MusicBrainz, if available. These images and data will be re-displayed from the cache if the track is replayed. Timestamps are also displayed on the traffic imagery to reflect how current it is.
    4. Resizable images: All images (logos, album art, traffic, and weather) will resize to fill up their respective areas. Likewise, their size will change as the window is resized.
    Source code(tar.gz)
    Source code(zip)
Owner
Retired Operations Manager trying to get back to developing neat radio stuff, daily 10ks, and supporting Arsenal. Oh, and I really need to update my 10+yo avi
null
Hiiii this is the Spanish for Linux and win 10 and in the near future the english version of PortScan my new tool on which you can see what ports are Open only with the IP adress.

PortScanner-by-IIT PortScanner es una herramienta programada en Python3. Como su nombre indica esta herramienta escanea los primeros 150 puertos de re

null 5 Sep 19, 2022
Handwriting Recognition System based on a deep Convolutional Recurrent Neural Network architecture

Handwriting Recognition System This repository is the Tensorflow implementation of the Handwriting Recognition System described in Handwriting Recogni

Edgard Chammas 346 Jan 7, 2023
This is the official PyTorch implementation of the paper "TransFG: A Transformer Architecture for Fine-grained Recognition" (Ju He, Jie-Neng Chen, Shuai Liu, Adam Kortylewski, Cheng Yang, Yutong Bai, Changhu Wang, Alan Yuille).

TransFG: A Transformer Architecture for Fine-grained Recognition Official PyTorch code for the paper: TransFG: A Transformer Architecture for Fine-gra

Ju He 307 Jan 3, 2023
Using computer vision method to recognize and calcutate the features of the architecture.

building-feature-recognition In this repository, we accomplished building feature recognition using traditional/dl-assisted computer vision method. Th

null 4 Aug 11, 2022
Just a script for detecting the lanes in any car game (not just gta 5) with specific resolution and road design ( very basic and limited )

GTA-5-Lane-detection Just a script for detecting the lanes in any car game (not just gta 5) with specific resolution and road design ( very basic and

Danciu Georgian 4 Aug 1, 2021
Developed an AI-based system to control the mouse cursor using Python and OpenCV with the real-time camera.

Developed an AI-based system to control the mouse cursor using Python and OpenCV with the real-time camera. Fingertip location is mapped to RGB images to control the mouse cursor.

Ravi Sharma 71 Dec 20, 2022
A document scanner application for laptops/desktops developed using python, Tkinter and OpenCV.

DcoumentScanner A document scanner application for laptops/desktops developed using python, Tkinter and OpenCV. Directly install the .exe file to inst

Harsh Vardhan Singh 1 Oct 29, 2021
A Screen Translator/OCR Translator made by using Python and Tesseract, the user interface are made using Tkinter. All code written in python.

About An OCR translator tool. Made by me by utilizing Tesseract, compiled to .exe using pyinstaller. I made this program to learn more about python. I

Fauzan F A 41 Dec 30, 2022
It is a image ocr tool using the Tesseract-OCR engine with the pytesseract package and has a GUI.

OCR-Tool It is a image ocr tool made in Python using the Tesseract-OCR engine with the pytesseract package and has a GUI. This is my second ever pytho

Khant Htet Aung 4 Jul 11, 2022
This is a GUI for scrapping PDFs with the help of optical character recognition making easier than ever to scrape PDFs.

pdf-scraper-with-ocr With this tool I am aiming to facilitate the work of those who need to scrape PDFs either by hand or using tools that doesn't imp

Jacobo José Guijarro Villalba 75 Oct 21, 2022
This is a GUI program which consist of 4 OpenCV projects

Tkinter-OpenCV Project Using Tkinter, Opencv, Mediapipe This is a python GUI program using Tkinter which consist of 4 OpenCV projects 1. Finger Counte

Arya Bagde 3 Feb 22, 2022
Image processing in Python

scikit-image: Image processing in Python Website (including documentation): https://scikit-image.org/ Mailing list: https://mail.python.org/mailman3/l

Image Processing Toolbox for SciPy 5.2k Dec 30, 2022
Packaged, Pytorch-based, easy to use, cross-platform version of the CRAFT text detector

CRAFT: Character-Region Awareness For Text detection Packaged, Pytorch-based, easy to use, cross-platform version of the CRAFT text detector | Paper |

null 188 Dec 28, 2022
A community-supported supercharged version of paperless: scan, index and archive all your physical documents

Paperless-ngx Paperless-ngx is a document management system that transforms your physical documents into a searchable online archive so you can keep,

null 5.2k Jan 4, 2023
An Implementation of the alogrithm in paper IncepText: A New Inception-Text Module with Deformable PSROI Pooling for Multi-Oriented Scene Text Detection

InceptText-Tensorflow An Implementation of the alogrithm in paper IncepText: A New Inception-Text Module with Deformable PSROI Pooling for Multi-Orien

GeorgeJoe 115 Dec 12, 2022
Repository collecting all the submodules for the new PyTorch-based OCR System.

OCRopus3 is being replaced by OCRopus4, which is a rewrite using PyTorch 1.7; release should be soonish. Please check github.com/tmbdev/ocropus for up

NVIDIA Research Projects 138 Dec 9, 2022
WACV 2022 Paper - Is An Image Worth Five Sentences? A New Look into Semantics for Image-Text Matching

Is An Image Worth Five Sentences? A New Look into Semantics for Image-Text Matching Code based on our WACV 2022 Accepted Paper: https://arxiv.org/pdf/

Andres 13 Dec 17, 2022