WhiteCane is a virtual assistant that helps the user in visualising his surroundings. It is accessible through a seamless voice based interface with no hard coded commands.

The query of the user is classified into various intents using Language Processing techniques and the image is accordingly fed to an AI model to get the output. This output is then spoken back to the user in a proper format.

Modes:

Currently, WhiteCane supports 3 modes, which are as follow.

1. Main mode

• Scene Description

• Currency Recognition

• OCR

• Help me Find

• Visual Question Answering

• Navigation

• Help.

2. People mode

• Count

• Emotion

• Gender

• Age.

3. Research Mode (Experimental)

• Title

• Authors

• Affiliations

• Email

• Abstract.

NOTE:

Many of the features are still under development and do not have 100% accuracy. Please take this into consideration while using this app.