Key_value Computer Vision Project

Shiva Reddy

Updated 2 years ago

25

views

1

download
Classes (3)
Key
Noise
Value
Description

Here are a few use cases for this project:

  1. Document Processing and Analysis: Use "Key_value" to parse forms, invoices, receipts, or any printed or digital documents that have a 'Key-Value' pair structure. The model will identify and separate the 'Keys', 'Values', and 'Noise'.

  2. Data Extraction: Use this computer vision model for extracting key information from business cards, IDs, or certificates, distinguishing valuable data from noise.

  3. Image-Based Search Engines: Implement "Key_value" to extract and interpret textual content within images, categorizing it as 'key', 'value', or 'noise'. These results can be used as tags to index and search images more accurately.

  4. Optical Character Recognition Improvement: Enhance OCR systems with "Key_value". It can be used to identify and segregate actual data from noise in scanned images, enabling clearer and more efficient text recognition.

  5. Automated Data Entry: Use "Key_value" in creating a system for automatic data entry where the model can extrapolate 'keys' and their corresponding 'values' from images of structured forms, reducing manual labor and errors.

Supervision

Build Computer Vision Applications Faster with Supervision

Visualize and process your model results with our reusable computer vision tools.

Cite This Project

LICENSE
CC BY 4.0

If you use this dataset in a research paper, please cite it using the following BibTeX:

                        @misc{
                            key_value-up00u_dataset,
                            title = { Key_value Dataset },
                            type = { Open Source Dataset },
                            author = { Shiva Reddy },
                            howpublished = { \url{ https://universe.roboflow.com/shiva-reddy/key_value-up00u } },
                            url = { https://universe.roboflow.com/shiva-reddy/key_value-up00u },
                            journal = { Roboflow Universe },
                            publisher = { Roboflow },
                            year = { 2022 },
                            month = { nov },
                            note = { visited on 2025-02-18 },
                            }
                        
                    

Similar Projects

See More