Anti-Cheat System

Users can simply use photos or pictures from the internet and not taken in real time to upload to the system. By doing that, their contributions to the community will be poor and not valuable. We strictly prohibit that.

Therefore, the anti-cheat system has been researched and developed by us from the very beginning. To prevent user frauds, we’re using a special algorithm, by combining mobile device sensors such as accelerometer, gps,... and deep learning model. Nearly every scenario aimed at spoofing is detected. If you try to take a picture of the place on a display, our model will quickly detect the image as a 2D plane. Our model closely combines with data from sensors and evaluates the realism of the image, even, the model will reject poor quality photos such as blurred, lack of content in case of users take pictures irresponsibly.

Our team is continuing to explore and visualize invalid scenarios to improve the app's anti-cheat system to ensure that user reviews of tourist attractions and products are valid and has real value.