Datasets:
				
			
			
	
			
			
	
		
		Update README.md
Browse files
    	
        README.md
    CHANGED
    
    | 
         @@ -44,13 +44,17 @@ size_categories: 
     | 
|
| 44 | 
         | 
| 45 | 
         
             
            A comprehensive dataset designed for building and evaluating **end-to-end signature analysis pipelines**, including **signature detection** in document images and **signature verification** using genuine/forged pair classification.
         
     | 
| 46 | 
         | 
| 47 | 
         
            -
            **Developed by**:[@Mels22](https://huggingface.co/Mels22) and [@JoeCao](https://huggingface.co/JoeCao)
         
     | 
| 48 | 
         | 
| 49 | 
         
             
            ## Pipeline Overview
         
     | 
| 50 | 
         | 
| 51 | 
         
             
            This dataset supports a complete **signature detection and verification pipeline**. The process involves identifying the signature in a document and comparing it with a reference to determine if it is genuine or forged.
         
     | 
| 52 | 
         | 
| 53 | 
         
            -
             
     | 
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 54 | 
         | 
| 55 | 
         
             
            - The **Detection Model** locates the signature in the document.
         
     | 
| 56 | 
         
             
            - The cropped signature (`to_verify_signature`) is passed along with a sample signature (`sample_signature`) to the **Verification Model**.
         
     | 
| 
         @@ -99,3 +103,16 @@ To create a seamless, unified pipeline dataset for **detection + verification**, 
     | 
|
| 99 | 
         
             
            This process enables researchers to train and evaluate models for **both signature localization and signature verification** in a realistic, document-centric setting.
         
     | 
| 100 | 
         | 
| 101 | 
         | 
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
|
| 
         | 
| 
         | 
|
| 44 | 
         | 
| 45 | 
         
             
            A comprehensive dataset designed for building and evaluating **end-to-end signature analysis pipelines**, including **signature detection** in document images and **signature verification** using genuine/forged pair classification.
         
     | 
| 46 | 
         | 
| 47 | 
         
            +
            **Developed by**: [@Mels22](https://huggingface.co/Mels22) and [@JoeCao](https://huggingface.co/JoeCao)
         
     | 
| 48 | 
         | 
| 49 | 
         
             
            ## Pipeline Overview
         
     | 
| 50 | 
         | 
| 51 | 
         
             
            This dataset supports a complete **signature detection and verification pipeline**. The process involves identifying the signature in a document and comparing it with a reference to determine if it is genuine or forged.
         
     | 
| 52 | 
         | 
| 53 | 
         
            +
            <div style="text-align: center;">
         
     | 
| 54 | 
         
            +
                <img src="pipeline.png" alt="Detection and Verification Pipeline" style="display: block; margin: auto;">
         
     | 
| 55 | 
         
            +
                <div style="font-style: italic;">Figure 1: Detection and Verification Pipeline.</div>
         
     | 
| 56 | 
         
            +
            </div>
         
     | 
| 57 | 
         
            +
            <br>
         
     | 
| 58 | 
         | 
| 59 | 
         
             
            - The **Detection Model** locates the signature in the document.
         
     | 
| 60 | 
         
             
            - The cropped signature (`to_verify_signature`) is passed along with a sample signature (`sample_signature`) to the **Verification Model**.
         
     | 
| 
         | 
|
| 103 | 
         
             
            This process enables researchers to train and evaluate models for **both signature localization and signature verification** in a realistic, document-centric setting.
         
     | 
| 104 | 
         | 
| 105 | 
         | 
| 106 | 
         
            +
            ## Sample Code
         
     | 
| 107 | 
         
            +
            ```python
         
     | 
| 108 | 
         
            +
            from datasets import load_dataset
         
     | 
| 109 | 
         
            +
            data = load_dataset("Mels22/SigDetectVerifyFlow")
         
     | 
| 110 | 
         
            +
             
     | 
| 111 | 
         
            +
            for i, example in enumerate(data['train']):
         
     | 
| 112 | 
         
            +
                example['document'].show()
         
     | 
| 113 | 
         
            +
                example['to_verify_signature'].show()
         
     | 
| 114 | 
         
            +
                example['sample_signature'].show()
         
     | 
| 115 | 
         
            +
                print(f"Bbox: {example['bbox']}")
         
     | 
| 116 | 
         
            +
                print(f"Label: {example['label']}")
         
     | 
| 117 | 
         
            +
                break
         
     | 
| 118 | 
         
            +
            ```
         
     |