# Mitigating the Hubness Problem for Zero-Shot Learning of 3D Objects

arXiv: https://arxiv.org/abs/1907.06371 · HTML: https://mlsanity.com/paper/1907.06371/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Jul 2019.

## Built on by (most cited first, 24 of 24)

- [OpenScene: 3D Scene Understanding with Open Vocabularies](https://mlsanity.com/paper/2211.15654/) (2022)
- [PointCLIP: Point Cloud Understanding by CLIP](https://mlsanity.com/paper/2112.02413/) (2021)
- [PointCLIP V2: Prompting CLIP and GPT for Powerful 3D Open-world Learning](https://mlsanity.com/paper/2211.11682/) (2022)
- [CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP](https://mlsanity.com/paper/2301.04926/) (2023)
- [Semantic-aware Knowledge Distillation for Few-Shot Class-Incremental Learning](https://mlsanity.com/paper/2103.04059/) (2021)
- [CLIP-FO3D: Learning Free Open-world 3D Scene Representations from 2D Dense CLIP](https://mlsanity.com/paper/2303.04748/) (2023)
- [Cross Modal Retrieval with Querybank Normalisation](https://mlsanity.com/paper/2112.12777/) (2021)
- [CLIP2: Contrastive Language-Image-Point Pretraining from Real-World Point Cloud Data](https://mlsanity.com/paper/2303.12417/) (2023)
- [Generative Zero-Shot Learning for Semantic Segmentation of 3D Point Clouds](https://mlsanity.com/paper/2108.06230/) (2021)
- [Zero-Shot Learning on 3D Point Cloud Objects and Beyond](https://mlsanity.com/paper/2104.04980/) (2021)
- [Open-Vocabulary 3D Detection via Image-level Class and Debiased Cross-modal Contrastive Learning](https://mlsanity.com/paper/2207.01987/) (2022)
- [OpenSight: A Simple Open-Vocabulary Framework for LiDAR-Based Object Detection](https://mlsanity.com/paper/2312.08876/) (2023)
- [See More and Know More: Zero-shot Point Cloud Segmentation via Multi-modal Visual Data](https://mlsanity.com/paper/2307.10782/) (2023)
- [3D Compositional Zero-shot Learning with DeCompositional Consensus](https://mlsanity.com/paper/2111.14673/) (2021)
- [Geometrically-driven Aggregation for Zero-shot 3D Point Cloud Understanding](https://mlsanity.com/paper/2312.02244/) (2023)
- [Open-Vocabulary 3D Semantic Segmentation with Text-to-Image Diffusion Models](https://mlsanity.com/paper/2407.13642/) (2024)
- [Zero-shot point cloud segmentation by transferring geometric primitives](https://mlsanity.com/paper/2210.09923/) (2022)
- [Transferring CLIP's Knowledge into Zero-Shot Point Cloud Semantic Segmentation](https://mlsanity.com/paper/2312.07221/) (2023)
- [Open-Pose 3D Zero-Shot Learning: Benchmark and Challenges](https://mlsanity.com/paper/2312.07039/) (2023)
- [3D Point Cloud Network Pruning: When Some Weights Do not Matter](https://mlsanity.com/paper/2408.14601/) (2024)
- [Prompt-guided Scene Generation for 3D Zero-Shot Learning](https://mlsanity.com/paper/2209.14690/) (2022)
- [Generalized Zero-Shot Learning for Point Cloud Segmentation with Evidence-Based Dynamic Calibration](https://mlsanity.com/paper/2509.08280/) (2025)
- [ChatGPT-guided Semantics for Zero-shot Learning](https://mlsanity.com/paper/2310.11657/) (2023)
- [Query-aware Hub Prototype Learning for Few-Shot 3D Point Cloud Semantic Segmentation](https://mlsanity.com/paper/2512.08253/) (2025)

## Abstract (arXiv)

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.