2021 IEEE/CVF International Conference on Computer Vision (ICCV) · 2021 · 199 citations · 26 references
ComputingComputer ScienceComputer VisionMachine LearningNatural Language ProcessingDeep LearningImage ClassificationMeta-learningRepresentation LearningScene UnderstandingSemantic SegmentationClassifier FusionFew-shot LearningFew-shot SegmentationFew-shot Semantic SegmentationInductive Inference
Few-shot semantic segmentation typically uses a CNN encoder, decoder, and classifier, but meta‑learning all three components is difficult when only a single support image is available. The authors aim to simplify the meta‑learning task by training only the classifier while keeping the encoder and decoder pre‑trained. They introduce a Classifier Weight Transformer that dynamically adapts the pre‑trained classifier’s weights to each query image in an inductive manner. Experiments on two benchmarks demonstrate that the method outperforms state‑of‑the‑art approaches by a large margin. Code is available at https://github.com/zhiheLu/CWT-for-FSS.
A few-shot semantic segmentation model is typically composed of a CNN encoder, a CNN decoder and a simple classifier (separating foreground and background pixels). Most existing methods meta-learn all three model components for fast adaptation to a new class. However, given that as few as a single support set image is available, effective model adaption of all three components to the new class is extremely challenging. In this work we propose to simplify the meta-learning task by focusing solely on the simplest component – the classifier, whilst leaving the en-coder and decoder to pre-training. We hypothesize that if we pretrain an off-the-shelf segmentation model over a set of diverse training classes with sufficient annotations, the encoder and decoder can capture rich discriminative features applicable for any unseen classes, rendering the sub-sequent meta-learning stage unnecessary. For the classifier meta-learning, we introduce a Classifier Weight Transformer (CWT) designed to dynamically adapt the support-set trained classifier’s weights to each query image in an inductive way. Extensive experiments on two standard bench-marks show that despite its simplicity, our method outperforms the state-of-the-art alternatives, often by a large margin. Code is available on https://github.com/zhiheLu/CWT-for-FSS.
26
Deep Residual Learning for Image Recognition
Kaiming He, Xiangyu Zhang, Shaoqing Ren et al. · 2016 · 222.1K citations · Full text
DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · 2023 · 76.3K citations · Full text
ImageNet Large Scale Visual Recognition Challenge
Olga Russakovsky, Jia Deng, Hao Su et al. · International Journal of Computer Vision · 2015 · 40.2K citations
Fully convolutional networks for semantic segmentation
Jonathan Long, Evan Shelhamer, Trevor Darrell · 2015 · 36.9K citations