Saliency prediction has achieved significant success in the past decade . However, it still remains challenging to predict saliency maps on images in new domains that lack sufficient data for data-hungry models… To solve this problem, we propose a few-shot transfer learning paradigm for saliency prediction . The proposed framework is gradient-based and model-agnostic . The code is publicly available at \url{https://://://github.com/luoyan407/n-reference}. The code is public at http://://www.genean-lan-research-group.org/saliency-prediction-predicting-models-a-model-transfer-learning-training-model framework. We conduct comprehensive experiments and ablation study on various source domain and target domain pairs.

Links: PDF - Abstract

Code :

None

Keywords : saliency - prediction - transfer - model - learning -

Leave a Reply

Your email address will not be published. Required fields are marked *

error

Enjoy this blog? Please spread the word :)