An interesting new approach!
"... Humans and computer vision systems both have what is known as foveal vision, which is used for scrutinizing highly detailed objects. Humans also possess peripheral vision, which is used to organize a broad, spatial scene. Typical computer vision approaches attempt to model foveal vision — which is how a machine recognizes objects — and tend to ignore peripheral vision ...
But foveal computer vision systems are vulnerable to adversarial noise, which is added to image data by an attacker. ...
The results suggest that designing a machine-learning model to include some form of peripheral processing could enable the model to automatically learn visual representations that are robust to some subtle manipulations in image data. ..."
But foveal computer vision systems are vulnerable to adversarial noise, which is added to image data by an attacker. ...
The results suggest that designing a machine-learning model to include some form of peripheral processing could enable the model to automatically learn visual representations that are robust to some subtle manipulations in image data. ..."
No comments:
Post a Comment