Research
Emergent Neural Network Mechanisms for Generalization to Objects in Novel Orientations
arXiv:2109.13445v3 Announce Type: replace-cross Abstract: The capability of Deep Neural Networks (DNNs) to recognize objects in orientations outside the distribution of the training data is not well u
arXiv:2109.13445v3 Announce Type: replace-cross Abstract: The capability of Deep Neural Networks (DNNs) to recognize objects in orientations outside the distribution of the training data is not well understood. We present evidence that DNNs are capable of generalizing to objects in novel orientations by disseminating orientation-invariance obtained from familiar objects seen from many viewpoints. This capability strengthens when training the DNN with an increasing number of familiar objects, but only in orientations that involve 2D rotations of familiar orientations. We show that this dissemination is achieved via neurons tuned to common features between familiar and unfamiliar objects. These results implicate brain-like neural mechanisms for generalization.
Related
- Towards Understanding Deep Learning Model in Image Recognition via Coverage Test
- Enhancing Deep Neural Network Reliability with Refinement and Calibration
- Circuit-Inspired High-Order Neural Networks with Unified Neural Dynamics Modeling for PDE Solving and Visual Perception
- Geometrically Constrained Outlier Synthesis
Source: arXiv cs.AI | 2026-08-12