Snake bites are one of the most deadliest neglected tropical disease causing over 100,000 deaths and 400,000 victims of disability and disfigurement globally every year.
Snake bites are one of the most deadliest neglected tropical disease causing over 100,000 deaths and 400,000 victims of disability and disfigurement globally every year.
...
@@ -9,29 +10,38 @@ The poor and rural communites are the worst affected due to their limited access
...
@@ -9,29 +10,38 @@ The poor and rural communites are the worst affected due to their limited access
In this challenge we explore the use of computer vision in correctly classifying snakes as `venomous` / `non venomous`.
In this challenge we explore the use of computer vision in correctly classifying snakes as `venomous` / `non venomous`.
</br>Understand with code! Here is [`getting started code`](https://discourse.aicrowd.com/t/baseline-mnist/2757) for you.`😄`
Understand with code! Here is [`getting started code`](https://discourse.aicrowd.com/t/baseline-mnist/2757) for you.`😄`
# 💾 Dataset
# 💾 Dataset
The dataset provided has many variations in it:
The dataset provided contains images of various species of snake with each image of size 224*224 pixels. The train set and validation set has around 54,000 and 6000 images respectively grouped into the classes of `venomous` and `non venomous` while the test set contains around 15,000 images which needs to be predicted. The dataset is quite vast and has following variations:
* variation by age
* variation by age
* variation by geography
* variation by geography
* medically important venomous snakes
* medically important venomous snakes
* similar looking harmless snakes
* similar looking harmless snakes
<br>
# 📁 Files
# 📁 Files
Following files are available in the `resources` section:
Following files are available in the `resources` section:
There are four files for this challenge.
There are four files for this challenge.
-`train.tar.gz` : contains approximately 54k training images.
-`train.zip` : (`54,000`) The zip file contains `train` folder with two folder inside it with name as `venomous` and `non venomous`.These folder contains the images of respective classes.
-`val.tar.gz` : contains approximately 6k validation images.
-`test.tar.gz` : contains approximately 15k test images.
-`val.zip` : (`6,000`) The zip file contains `val` folder with two folder inside it with name as `venomous` and `non venomous`.These folder contains the images of respective classes.
-`test.zip` : (`15,000`) The zip file contains `test` folder with images of the snakes to whose classes needs to be predicted.
-`sample_submission.csv` : Sample submission format for the challenge.
# 🚀 Submission
# 🚀 Submission
- Prepare a CSV containing header as [header] and predicted value as digit 0 or 1 respectively denoting whether or not the corresponding ad will be clicked upon.
- Prepare a CSV containing two coloums with header `id` and `class`. The first column contains the image id(without the ".jpg" extension) and the second as the class it belongs to `venomous` and `non_venomous`.
- Name of the above file should be submission.csv.
- Name of the above file should be submission.csv.
- Sample submission format available at sample_submission.csv in the resorces section.
- Sample submission format available at sample_submission.csv in the resorces section.
...
@@ -43,8 +53,7 @@ During evaluation [F1 score](https://scikit-learn.org/stable/modules/generated/s
...
@@ -43,8 +53,7 @@ During evaluation [F1 score](https://scikit-learn.org/stable/modules/generated/s
