Spaces:

Jay9115
/

Deep_fake_Model_load

Sleeping

File size: 12,726 Bytes

eeeea78

"""
Postman Testing Guide for GradCAM API
Complete instructions for testing all GradCAM endpoints
"""

# ==============================================================================
# POSTMAN SETUP AND TESTING GUIDE FOR GRADCAM API
# ==============================================================================

POSTMAN_COLLECTION_JSON = '''
{
  "info": {
    "name": "GradCAM Deepfake Detection API",
    "description": "Test collection for GradCAM visualization endpoints",
    "version": "1.0.0"
  },
  "variable": [
    {
      "key": "base_url",
      "value": "http://localhost:5000",
      "type": "string"
    }
  ],
  "item": [
    {
      "name": "1. Test Service Status",
      "request": {
        "method": "GET",
        "header": [],
        "url": {
          "raw": "{{base_url}}/api/gradcam/test",
          "host": ["{{base_url}}"],
          "path": ["api", "gradcam", "test"]
        },
        "description": "Test if GradCAM service is running and model is loaded"
      }
    },
    {
      "name": "2. Get Model Information",
      "request": {
        "method": "GET", 
        "header": [],
        "url": {
          "raw": "{{base_url}}/api/gradcam/model-info",
          "host": ["{{base_url}}"],
          "path": ["api", "gradcam", "model-info"]
        },
        "description": "Get detailed information about GradCAM capabilities and available layers"
      }
    },
    {
      "name": "3. Single Image Analysis",
      "request": {
        "method": "POST",
        "header": [],
        "body": {
          "mode": "formdata",
          "formdata": [
            {
              "key": "image",
              "type": "file",
              "src": ""
            },
            {
              "key": "return_image",
              "value": "true",
              "type": "text"
            },
            {
              "key": "layer_name",
              "value": "",
              "type": "text"
            }
          ]
        },
        "url": {
          "raw": "{{base_url}}/api/gradcam/analyze",
          "host": ["{{base_url}}"],
          "path": ["api", "gradcam", "analyze"]
        },
        "description": "Analyze single image with GradCAM visualization"
      }
    },
    {
      "name": "4. Batch Image Analysis",
      "request": {
        "method": "POST",
        "header": [],
        "body": {
          "mode": "formdata",
          "formdata": [
            {
              "key": "images",
              "type": "file",
              "src": ""
            },
            {
              "key": "images",
              "type": "file", 
              "src": ""
            },
            {
              "key": "max_images",
              "value": "5",
              "type": "text"
            }
          ]
        },
        "url": {
          "raw": "{{base_url}}/api/gradcam/analyze-batch",
          "host": ["{{base_url}}"],
          "path": ["api", "gradcam", "analyze-batch"]
        },
        "description": "Batch analysis of multiple images"
      }
    },
    {
      "name": "5. Backend Status Check",
      "request": {
        "method": "GET",
        "header": [],
        "url": {
          "raw": "{{base_url}}/",
          "host": ["{{base_url}}"],
          "path": [""]
        },
        "description": "Check overall backend status and available endpoints"
      }
    }
  ]
}
'''

# ==============================================================================
# STEP-BY-STEP TESTING INSTRUCTIONS
# ==============================================================================

TESTING_INSTRUCTIONS = """
📋 POSTMAN TESTING GUIDE - GradCAM API
====================================

🚀 SETUP
--------
1. Start your Flask backend:
   cd "d:\\SGP\\DeepFake\\Deepfake-Detection\\Webapp\\Model_Load"
   python app.py

2. Backend should be running on: http://localhost:5000

3. Import Postman Collection:
   - Copy the JSON collection above
   - In Postman: Import > Raw Text > Paste JSON > Import

📝 TEST SEQUENCE
---------------

TEST 1: Service Status Check
---------------------------
• Method: GET
• URL: http://localhost:5000/api/gradcam/test
• Headers: None required
• Body: None

Expected Response:
{
  "status": "GradCAM Analysis Service Active",
  "model_loaded": true,
  "face_detector_ready": true,
  "memory_optimization": "Enabled",
  "capabilities": [...],
  "test_successful": true
}

✅ Success: Status 200, test_successful = true
❌ Fail: Status 500, check model loading


TEST 2: Model Information
------------------------
• Method: GET  
• URL: http://localhost:5000/api/gradcam/model-info
• Headers: None required
• Body: None

Expected Response:
{
  "model_name": "XceptionNet with GradCAM",
  "description": "Visual explanation of deepfake detection...",
  "capabilities": {...},
  "available_layers": [...],
  "recommended_layers": [...]
}

✅ Success: Status 200, layers list populated
❌ Fail: Status 500, model not loaded


TEST 3: Single Image Analysis  
-----------------------------
• Method: POST
• URL: http://localhost:5000/api/gradcam/analyze
• Headers: None required (auto-detected)

Body (form-data):
┌─────────────────┬──────────┬─────────────────────────────────┐
│ Key             │ Type     │ Value                           │
├─────────────────┼──────────┼─────────────────────────────────┤
│ image           │ file     │ [Select image with face]        │
│ return_image    │ text     │ true                            │
│ layer_name      │ text     │ (leave empty for auto-detect)   │
└─────────────────┴──────────┴─────────────────────────────────┘

Expected Response:
{
  "success": true,
  "prediction": {
    "classification": "DEEPFAKE" or "AUTHENTIC",
    "is_deepfake": true/false,
    "confidence": 0.85,
    "raw_score": 0.7234
  },
  "face_detection": {
    "face_found": true,
    "face_confidence": 0.99,
    "face_coordinates": {...}
  },
  "gradcam_analysis": {
    "layer_used": "block14_sepconv1_act",
    "heatmap_statistics": {...}
  },
  "visualization": {
    "format": "PNG (base64 encoded)",
    "data": "iVBORw0KGgoAAAANSUhEUgAA...",
    "description": "GradCAM visualization..."
  },
  "interpretation": {...},
  "metadata": {...}
}

✅ Success: Status 200, success = true, visualization data present
❌ Fail: Status 400, no face detected or invalid image


TEST 4: Batch Analysis
---------------------
• Method: POST
• URL: http://localhost:5000/api/gradcam/analyze-batch
• Headers: None required

Body (form-data):
┌─────────────────┬──────────┬─────────────────────────────────┐
│ Key             │ Type     │ Value                           │
├─────────────────┼──────────┼─────────────────────────────────┤
│ images          │ file     │ [Select first image]            │
│ images          │ file     │ [Select second image]           │
│ images          │ file     │ [Select more images...]         │
│ max_images      │ text     │ 5                               │
└─────────────────┴──────────┴─────────────────────────────────┘

Note: Add multiple 'images' keys for multiple files

Expected Response:
{
  "success": true,
  "batch_summary": {
    "total_files": 3,
    "processed_successfully": 2,
    "failed_files": 1,
    "deepfake_detected": 1,
    "authentic_detected": 1,
    "average_confidence": 0.78
  },
  "results": [...],
  "failed_files": [...],
  "metadata": {...}
}

✅ Success: Status 200, batch_summary populated
❌ Fail: Status 400, no valid images provided


TEST 5: Backend Compatibility
-----------------------------
• Method: GET
• URL: http://localhost:5000/
• Headers: None required
• Body: None

Expected Response:
{
  "status": "ok",
  "message": "Deepfake Detection API Backend is running",
  "available_endpoints": {
    "gradcam_analysis": "/api/gradcam/analyze",
    "gradcam_batch": "/api/gradcam/analyze-batch",
    "gradcam_info": "/api/gradcam/model-info",
    "gradcam_test": "/api/gradcam/test"
  },
  "new_features": {
    "gradcam_visualization": "Visual explanation of AI decisions"
  }
}

✅ Success: Status 200, gradcam endpoints listed
❌ Fail: Backend not running or integration failed
"""

# ==============================================================================
# COMMON ERROR CODES AND SOLUTIONS
# ==============================================================================

ERROR_TROUBLESHOOTING = """
🔧 TROUBLESHOOTING GUIDE
========================

ERROR 1: Connection Refused
---------------------------
Symptoms: Can't connect to localhost:5000
Solutions:
• Ensure Flask app is running: python app.py
• Check if port 5000 is available
• Verify no firewall blocking
• Try: curl http://localhost:5000/

ERROR 2: Model Not Loaded (Status 500)
--------------------------------------  
Symptoms: test_successful = false, model_loaded = false
Solutions:
• Check if xceptionnet.keras exists in models/ folder
• Verify TensorFlow installation: pip install tensorflow
• Check file permissions on model file
• Review server logs for detailed error

ERROR 3: No Face Detected (Status 400)
--------------------------------------
Symptoms: "No face detected" error in analysis
Solutions:
• Use images with clear, visible faces
• Ensure face is well-lit and facing camera
• Try different images with larger faces
• Check MTCNN installation: pip install mtcnn

ERROR 4: Memory Error (Status 500)
----------------------------------
Symptoms: Out of memory during processing
Solutions:
• Reduce image size before upload (< 2MB recommended)
• Reduce batch size (max_images parameter)
• Restart Flask server to clear memory
• Check available system RAM

ERROR 5: Invalid Image Format
-----------------------------
Symptoms: "Invalid image file" error
Solutions:
• Use supported formats: JPG, JPEG, PNG
• Ensure file is not corrupted
• Check file size (< 10MB recommended)
• Try re-saving image in different format

ERROR 6: Import Errors (Status 500)
-----------------------------------
Symptoms: Module import failures in logs
Solutions:
• Install missing dependencies: pip install -r requirements.txt
• Check virtual environment activation
• Verify Python version compatibility (3.7+)
• Update packages: pip install --upgrade tensorflow opencv-python

PERFORMANCE TIPS:
• Use images 800x600 or smaller for faster processing
• JPEG format is typically faster than PNG
• Batch processing is more efficient for multiple images
• Clear browser cache if using web interface
"""

# ==============================================================================
# SAMPLE CURL COMMANDS (Alternative to Postman)
# ==============================================================================

CURL_EXAMPLES = '''
🌐 CURL COMMAND EXAMPLES
========================

# Test 1: Service Status
curl -X GET http://localhost:5000/api/gradcam/test

# Test 2: Model Info  
curl -X GET http://localhost:5000/api/gradcam/model-info

# Test 3: Single Image Analysis
curl -X POST http://localhost:5000/api/gradcam/analyze \\
  -F "image=@/path/to/your/image.jpg" \\
  -F "return_image=true" \\
  -F "layer_name="

# Test 4: Batch Analysis
curl -X POST http://localhost:5000/api/gradcam/analyze-batch \\
  -F "images=@/path/to/image1.jpg" \\
  -F "images=@/path/to/image2.jpg" \\
  -F "max_images=5"

# Test 5: Backend Status
curl -X GET http://localhost:5000/
'''

if __name__ == "__main__":
    print("📖 POSTMAN TESTING GUIDE FOR GRADCAM API")
    print("="*50)
    
    # Save collection to file
    with open("GradCAM_API_Collection.json", "w") as f:
        f.write(POSTMAN_COLLECTION_JSON)
    
    print("✅ Postman collection saved to: GradCAM_API_Collection.json")
    print("\n📋 TESTING INSTRUCTIONS:")
    print(TESTING_INSTRUCTIONS)
    print("\n🔧 TROUBLESHOOTING:")
    print(ERROR_TROUBLESHOOTING)
    print("\n🌐 CURL ALTERNATIVES:")
    print(CURL_EXAMPLES)