Generating natural language description of an image, also known as image captioning is a well-studied problem. The task of image captioning requires detection of various objects and their attributes in the image, the relation between these objects and generating a syntactic and semantic description in natural language description. Image captioning is important for multiple applications such as Content Based Image Retrieval and social media websites for recommending automatic captions of images.
Prizes worth INR 30000
Test Link: https://www.hackerearth.com/challenges/college/pragyan-19-ml-challenge/
Short Link: http://hck.re/p19mc1
- This event contains a single round which will be conducted over two days.
- You will be provided with the dataset of a company. Using this, you will have to make a program that predicts the label for the images provided.
- You have to submit the program in a file with .csv extension.
-
File description:
3 files would be provided to you.
- Training.csv → The training set
- Test.csv → The test set
- Sample Submission.csv → A sample submission file in the correct format.
Rules:
-
The languages allowed are Python and R.
-
Teams should work independently. Sharing answers or collaborating with another team to secure multiple spots at finals constitutes cheating.
-
Only queries regarding technical issues will be entertained.
-
The judges’ decision will be final.
Judging criteria:
Bleu metric
Problem Statement will be updated when the challenge begins.
F.A.Q:
Q) In case of a tie, what will be the criteria for judgement?
A) In case of tie in Bleu metric, time of submission will be taken into account.
Contacts:
-
Sanjeeda Khatoon: +91 87891 87123