This repository contains code and resources for performing zero-shot object detection and segmentation using deep learning techniques. Zero-shot learning enables the model to detect and segment objects from classes that were not seen during training, by leveraging semantic information and transferring knowledge from seen classes.
The AWS CloudFormation Stack can be created using 2 methods: (1) Using Template or (2) Using AWS CDK. Both the methods are described as follows:
-
[WIP] Create Stack using AWS CloudFormation:
-
Using AWS CLI:
- If AWS CLI is installed, use it to create the CloudFormation Stack:
$ git clone https://github.com/aws-samples/zero-shot-learning-object-detection-and-segmentation $ cd zero-shot-learning-object-detection-and-segmentation $ aws cloudformation create-stack \ --stack-name ZSL-STACK \ --template-body file://zsl-cdk.yaml \ --capabilities CAPABILITY_IAM
- Once the Stack is created, navigate to SageMaker in AWS Console
- From SageMaker -> Notebooks -> Open ZSL Notebook
- Follow the steps in the Notebook
See CONTRIBUTING for more information.
This library is licensed under the MIT-0 License. See the LICENSE file.
- Fabian Benitez-Quiroz
- Junjie Tang
- Romil Shah