We use our in-house built deep learning model that involves multiple layers of processing. The model is capable of examining images from multiple different perspectives and dimensions. Through some mathematical magic, our network becomes more accurate over time.
The Frameright AI runs in the background of the Frameright interface, suggesting suitable crops. You can then adjust them as needed, and the AI will follow your direction in real-time.
Our "augmented intelligence" approach combines the speed of machines with the accuracy of humans, allowing for immediate corrections to mistakes in the workflow. This means that errors in image crops caused by the AI can be corrected right away, resulting in a smoother and more efficient workflow. The model also learns from the changes you make and evolves to crop images in a way that matches your personal preferences and style.