# Bleeding Detection and Classification

 ### Classification Metrics

| Metric (in probability)| Value    |

|------------------------|----------|

| Accuracy               |   0.7864 |

| Recall (Bleeding)                 |   0.9602 |

| Recall (Non Bleeding)                 |   0.6126|

| Recall (Weighted)                 |   0.7864|

| F1-Score (Bleeding)               |  0.8180  |

| F1-Score (Non Bleeding)               |  0.7415  |

| F1-Score (Weighted)               |  0.7798  |

### Detection Metrics

| Metric (in probability)| Value          |

|------------------------|----------------|

|  mAP50 |     0.652      |

| mAP50-95|     0.336     |

### Training Metrics

| **Confusion Metrics** | **Confusion Metrics Normalized**|

|------ |---------------------|

|![confusion_matrix](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/31ef191b-b8f4-4dc8-a336-6a2a340db6d3)|![confusion_matrix_normalized](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/7a7d65d3-aa1c-43c1-bf19-7c40d390da9f)|

| **F1 Curve** | **P Curve**|**PR Curve** |**R Curve**|

|------ |---------------------|------ |---------------------|

|![F1_curve](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/ec099737-cbbb-4da3-9cfd-4a0bb5913a58)|![P_curve](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/a14ef28c-51e8-4c99-8ffd-58307e6c2280)|![PR_curve](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/e8073211-0c11-4e22-9559-7e50b6aa47da)|![R_curve](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/28c4cf9f-9e21-471a-92d8-de5b5ad80851)

## Detection and Classification

### Validation Dataset

| **Imagename** | **img- (43).png** | **img- (44).png**|**img- (68).png**|**img- (164).png**|**img- (200).png**|

|------ |---------------------|---------------------|---------------------|---------------------|---------------------|

|**Images** |![img- (43)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/dd721c32-092d-4c25-8cfc-d3b3e94c33de)|![img- (44)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/1dec6675-b3bc-4217-9e57-d237294a9142)|![img- (68)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/a9708925-b2a0-4156-9efd-5cc35a7f5a44)|![img- (164)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/4da25df9-a9c1-4512-af41-853264f08460)|![img- (200)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/d6217716-352c-49e6-a7a7-ca35a373852f)

| **Imagename** | **img- (448).png** | **img- (514).png**|**img- (628).png**|**img- (687).png**|**img- (677).png**|

|**Images** |![img- (448)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/a88865f2-2680-4226-976c-fe3c31813f0e)|![img- (514)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/624112c2-0168-4eaa-8282-ed38982dc949)|![img- (628)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/c44170d6-e515-48df-b284-d0c76fc632b9)|![img- (687)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/ba02798a-aa2c-4690-83ce-abf4feb964f4)|![img- (677)](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/c2974286-3a16-4b96-a895-c574ebd0b404)

## Interpretability Plots (Cam Plots of 2nd last layer)

### Test Dataset 1

| **Imagename** | **A0010.png** | **A0021.png**|**A0031.png**|**A0040.png**|**A0045.png**|

|------ |---------------------|---------------------|---------------------|---------------------|---------------------|

|**Cam Images** | ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/d4f9ba16-fa0a-477f-b65c-473bb858a8df)| ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/ead3de41-9dfd-4654-8a8b-dc264198a273)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/d0369a27-069f-41d7-b110-dbf56b24c5a4)| ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/92306a7b-edb8-4943-9f68-956665a44cb1)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/52080bf0-c102-4d83-b92b-75f4a6cb0b4e)

### Test Dataset 2

| **Imagename** | **A0164.png** | **A0256.png**|**A0300.png**|**A0384.png**|**A0473.png**|

|------ |---------------------|---------------------|---------------------|---------------------|---------------------|

|**CAM Images** | ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/80a92f3f-d64a-41d0-b16b-22723c67c7e1)| ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/37361d83-a8bd-4c8d-869e-a7cdbe11b89f)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/06f7f812-1a4d-4147-bd3a-274cb84b7851)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/c24f688e-c112-4249-8c2d-133a80c7a411)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/ce9addc6-25c2-4e1f-a5ce-52360d8b795b)

## Detection and Classification

### Testing Dataset 1

| **Imagename** | **A0047.png** | **A0043.png**|**A0048.png**|**A0040.png**|**A0035.png**|

|------ |---------------------|---------------------|---------------------|---------------------|---------------------|

|**Images** | ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/b81cf6fd-58eb-4812-a174-46d1af28ae72)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/4e51a286-98e9-44af-85b3-cc0cc79dfbef)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/fb8d5cc0-9d46-46b1-b9a7-6d46215ceb57)| ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/bd986059-75c4-40ef-be4f-1c334fb14766)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/b1a4b811-3341-41fc-bdda-c82a55c6641e)

### Testing Dataset 2

| **Imagename** | **A0258.png** | **A0260.png**|**A0500.png**|**A0261.png**|**A0498.png**|

|------ |---------------------|---------------------|---------------------|---------------------|---------------------|

|**Images** | ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/b3024ef7-79cf-46ed-b2c0-24962f9f0fdc)| ![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/43632a5e-2389-4648-9839-ead3c0abe104)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/e6d69155-ff8c-4064-8b03-ed3462270385)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/5ed99275-804b-41ac-8697-d53105f52676)|![image](https://github.com/pragneshbarik/misahub-challenge/assets/65221256/f38d6335-3211-427e-8cb3-83c146813db3)|

# Instructions

- Clone the repo

- Install requirements.txt via `pip install -r requirements.txt`

- Download and extract the model weights from [here](https://drive.google.com/drive/folders/15JUQSBSc9LGDM7J7Hhs053oWCTjpFqJx?usp=sharing)

- Do necessary changes in `test.py` file.

- Running the `test.py` generates the `predictions.csv` file