Inspiration

The motivation behind our idea stems from the realization that companies, regardless of size, are losing a considerable amount of money and customer trust due to piracy and copyright violations. For instance, a study on cord-cutting suggests that Netflix may suffer a monthly loss of $192 million from piracy as reported in following article . Given the advancements in the field of machine learning and AI, we developed a method to identify and tackle copyright and piracy violators.

What it does

The technique of embedding secret data, commonly referred to as steganography, presents a chance for detection of violators. This method allows for strong encoding and decoding of data hidden within digital format such as streams, movies, and images, as demonstrated in our project and inspired by the following work.

Technical details

A spatial transformer network is utilized to record and rectify the encoded image, making it resilient against slight shifts in perspective. To produce a final output with the same length as the original message, the modified image is processed through a combination of dense and convolutional layers, followed by a sigmoid activation. This results in precise encoding and decoding of the data.

Challenges

At present, our detection capabilities are limited to "watermarks" on images due to the complexity of the concept. However, with more time and resources, it could be expanded to include encoding of videos as well.

Built With

Share this project:

Updates