Adaptive Masked Proxies for Few-Shot Segmentation

Abstract

Deep learning has thrived by training on large-scale datasets. However, in robotics applications sample efficiency is critical. We propose a novel adaptive masked proxies method that constructs the final segmentation layer weights from few labelled samples. It utilizes multi-resolution average pooling on base embeddings masked with the label to act as a positive proxy for the new class, while fusing it with the previously learned class signatures. Our method is evaluated on PASCAL-$5^i$ dataset and outperforms the state-of-the-art in the few-shot semantic segmentation. Unlike previous methods, our approach does not require a second branch to estimate parameters or prototypes, which enables it to be used with 2-stream motion and appearance based segmentation networks. We further propose a novel setup for evaluating continual learning of object segmentation which we name incremental PASCAL (iPASCAL) where our method outperforms the baseline method. Our code is publicly available at https://github.com/MSiam/AdaptiveMaskedProxies.

arXiv commentAccepted to ICCV'19

References in corpus (13)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• High-Speed Tracking with Kernelized Correlation Filters
• Prototypical Networks for Few-shot Learning
• Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials
• Three scenarios for continual learning
• Wider or Deeper: Revisiting the ResNet Model for Visual Recognition
• FusionSeg: Learning to combine motion and appearance for fully automatic segmention of generic objects in videos
• SG-One: Similarity Guidance Network for One-Shot Semantic Segmentation
• Learning Video Object Segmentation with Visual Memory
• Low-Shot Learning with Imprinted Weights
• Transfer of View-manifold Learning to Similarity Perception of Novel Objects
• Learning Motion Patterns in Videos

Cited by in corpus (5)

• FECANet: Boosting Few-Shot Semantic Segmentation with Feature-Enhanced Context-Aware Network
• BriNet: Towards Bridging the Intra-class and Inter-class Gaps in One-Shot Segmentation
• CRNet: Cross-Reference Networks for Few-Shot Segmentation
• Prototype-based Incremental Few-Shot Semantic Segmentation
• Incremental Meta-Learning via Indirect Discriminant Alignment