Conditional Gaussian Distribution Learning for Open Set Recognition

By Xin Sun, Zhen Yang, Chi Zhang, Guohao Peng, Keck Voon Ling. 2020

274 citations recorded by the source corpus.

Deep neural networks have achieved state-of-the-art performance in a wide range of recognition/classification tasks. However, when applying deep learning to real-world applications, there are still multiple challenges. A typical challenge is that unknown samples may be fed into the system during the testing phase and traditional deep neural networks will wrongly recognize the unknown sample as one of the known classes. Open set recognition is a potential solution to overcome this problem, where the open set classifier should have the ability to reject unknown samples as well as maintain high classification accuracy on known classes. The variational auto-encoder (VAE) is a popular model to detect unknowns, but it cannot provide discriminative representations for known classification. In this paper, we propose a novel method, Conditional Gaussian Distribution Learning (CGDL), for open set recognition. In addition to detecting unknown samples, this method can also classify known samples by forcing different latent features to approximate different Gaussian models. Meanwhile, to avoid information hidden in the input vanishing in the middle layers, we also adopt the probabilistic ladder architecture to extract high-level abstract features. Experiments on several standard image datasets reveal that the proposed method significantly outperforms the baseline method and achieves new state-of-the-art results.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Han-Jia Ye
2. Ziwei Liu
3. De-chuan Zhan
4. Da-Wei Zhou
5. Guosheng Lin
6. Kaiyang Zhou
7. Andrew Zisserman
8. Chi Zhang

Citations in the corpus

• Learning Placeholders for Open-Set Recognition (2021, methodology): “The calibrated logits is [Wφ(x), ŵφ(x) + bias], we obtain the best bias by ensuring 95% of validation data in Dval to be recognized as known [29, 26].”
• Generalized Out-of-Distribution Detection: A Survey (2021, background): “Subsequent works use conditional Gaussian distributions by forcing different latent features to approximate class-wise Gaussian models, which enables classifying known samples as well as reject unknown samples [184].”
• Weakly Supervised Segmentation with Maximum Bipartite Graph Matching (2020, background): “In the past few years, Deep Neural Networks have made remarkable breakthroughs in many computer vision tasks, such as image classification [9, 49, 63], object detection [15] and semantic segmentation [7, 34, 35, 37, 64, 65].”
• Generalized Category Discovery (2022, methodology): “Several methods have been proposed to train models such that images with large reconstruction error are regarded as open-set samples [34, 41, 48].”
• CMG: A Class-Mixed Generation Approach to Out-of-Distribution Detection (2022, methodology): “Following [51] Setting 2 - Far-OOD Detection on Different Datasets: The IND data and OOD data come from different datasets.”
• Learning Bounds for Open-Set Learning (2021, background): “Under this assumption, the standard supervised learning is also regarded as closed-set learning (CSL) (Geng et al., 2018; Yang et al., 2020).”
• CRNet: Cross-Reference Networks for Few-Shot Segmentation (2020, background): “The task of few-shot learning is also related to open set problem [20], where the goal is only to detect data from novel classes.”