Mitigating the Hubness Problem for Zero-Shot Learning of 3D Objects
By Ali Cheraghian, Shafin Rahman, Dylan Campbell, Lars Petersson. 2019
48 citations recorded by the source corpus.
The development of advanced 3D sensors has enabled many objects to be captured in the wild at a large scale, and a 3D object recognition system may therefore encounter many objects for which the system has received no training. Zero-Shot Learning (ZSL) approaches can assist such systems in recognizing previously unseen objects. Applying ZSL to 3D point cloud objects is an emerging topic in the area of 3D vision, however, a significant problem that ZSL often suffers from is the so-called hubness problem, which is when a model is biased to predict only a few particular labels for most of the test instances. We observe that this hubness problem is even more severe for 3D recognition than for 2D recognition. One reason for this is that in 2D one can use pre-trained networks trained on large datasets like ImageNet, which produces high-quality features. However, in the 3D case there are no such large-scale, labelled datasets available for pre-training which means that the extracted 3D features are of poorer quality which, in turn, exacerbates the hubness problem. In this paper, we therefore propose a loss to specifically address the hubness problem. Our proposed method is effective for both Zero-Shot and Generalized Zero-Shot Learning, and we perform extensive evaluations on the challenging datasets ModelNet40, ModelNet10, McGill and SHREC2015. A new state-of-the-art result for both zero-shot tasks in the 3D case is established.
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- Renrui Zhang
- Ali Cheraghian
- Peng Gao
- Songyou Peng
- Christian Simon
- Shafin Rahman
- Hongsheng Li
- Ziyu Guo
Citations in the corpus
- Semantic-aware Knowledge Distillation for Few-Shot Class-Incremental Learning (2021, background): “Being inspired by the literature on zero-shot learning [7, 6, 5, 20, 37, 24, 25], given an image as input, we estimate the semantic word vectors for the input instead of directly predicting its class label.”
- Synthesized Feature based Few-Shot Class-Incremental Learning on a Mixture of Subspaces (2021, background): “Learning without forgetting using word vectors: Word vectors have shown promising success on various computer vision tasks such as zero-shot learning, few-shot learning, image/video captioning and visual question answering [15, 29, 39, 43, 9, 46, 47, 6, 5, 7].”
- PointCLIP: Point Cloud Understanding by CLIP (2021, background): “Based on this prior work, [5] further mitigates the hubness problem [62] resulted from low-quality extracted 3D features and [6] introduces a triplet loss for better performance in transductive settings, which allows to utilize unlabeled “unseen” data at training time.”
- Transductive Zero-Shot Learning for 3D Point Cloud Classification (2019, methodology): “For the experiments on 3D data, we compare with two 3D ZSL methods, ZSLPC [9] and MHPC [8], and three 2D ZSL methods, f-CLSWGAN [58], CADA-VAE [43], and QFSL [49].”
- OpenScene: 3D Scene Understanding with Open Vocabularies (2022, background): “A handful of works [7–10] attempt to address the 3D point classification and generation tasks.”