Search Results for author:
Found 4 papers, 2 papers with code
SpaceEditing: Integrating Human Knowledge into Deep Neural Networks via Interactive Latent Space Editing
We propose an interactive editing method that allows humans to help deep neural networks (DNNs) learn a latent space more consistent with human knowledge, thereby improving classification accuracy on indistinguishable ambiguous data.
SGDraw: Scene Graph Drawing Interface Using Object-Oriented Representation
However, it is difficult to draw a proper scene graph for image retrieval, image generation, and multi-modal applications.
Efficient Human-in-the-loop System for Guiding DNNs Attention
Attention guidance is an approach to addressing dataset bias in deep learning, where the model relies on incorrect features to make decisions.
DAMix: A Density-Aware Mixup Augmentation for Single Image Dehazing under Domain Shift
However, these methods are often subject to performance degradation when domain shifts are confronted.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
