Open Set Learning with Counterfactual Images

By Lawrence Neal, Matthew Lyle Olson, Xiaoli Z. Fern, Weng-Keen Wong, Fuxin Li. 2018

537 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Media Technology

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

In Media Technology

Citations in the corpus

Counterfactuals in Explainable Artificial Intelligence (XAI): Evidence from Human Reasoning (2019, background): “Counterfactuals are also used in generative adversarial networks (GANs) to improve training [Neal et al., 2018].”
Learning Placeholders for Open-Set Recognition (2021, methodology): “Traditional generative-based open-set models tend to utilize powerful generative models to mimic novel patterns [22, 26].”
Recent Advances in Open Set Recognition: A Survey (2018, methodology): “As most of them followed the evaluation protocol8 defined in [89] and did not provide source codes, similar to [3], [147], we here only compare with their published results.”
OpenGAN: Open-Set Recognition via Open Data Generation (2021, unknown): “MNIST/CIFAR/SVHN/TinyImageNet are widely used in the open-set literature, and we follow the literature to experiment with these datasets [33, 35].”
Learning Bounds for Open-Set Learning (2021, methodology): “The area under the receiver operating characteristic (AUROC) (Neal et al., 2018) is also frequently used (Chen et al.”
Zero-Shot Out-of-Distribution Detection Based on the Pre-trained Model CLIP (2021, methodology): “All these baselines use a CNN encoder architecture introduced in (Neal et al. 2018).”
Generalized Out-of-Distribution Detection: A Survey (2021, unknown): “3: Unknown Generation [160], [161], [162], [163], [164], [165], [166]”
CMG: A Class-Mixed Generation Approach to Out-of-Distribution Detection (2022, methodology): “So far, many approaches have been proposed to solve this problem, from distance-based meth-ods [2, 3, 12, 20], to generative models [60, 38, 41, 36] and self-supervised learning methods [4, 11, 17, 19].”
Conditional Gaussian Distribution Learning for Open Set Recognition (2020, methodology): “Following the protocol defined in [22], we use four standard image datasets: MNIST [18], SVHN [23], CIFAR-10 [15] and Tiny-ImageNet [17], to measure the model’s ability to identify unknown samples.”
Learning and the Unknown: Surveying Steps toward Open World Recognition (2019, methodology): “The work in this area has received hundreds of citations from a wide range of application areas that need, or are using, OSR including: • Audio analysis (Battaglino, Lepauloux, and (Chiachia et al. 2014; Pinto et al. 2015; Rattani, Scheirer, and Ross 2015; Juefei-Xu and Savvides 2016; G¨unther et al”
Generative-Discriminative Feature Representations for Open-Set Recognition (2020, methodology): “We use the network architecture proposed in [16].”
Generalized Category Discovery (2022, background): “GANs are often employed to generate adversarial samples to train an open-set classiﬁer, e.g, [14, 25, 32].”
C2AE: Class Conditioned Auto-Encoder for Open-Set Recognition (2019, methodology): “For experiments with the TinyImageNet [20], 20 1: AUROC for open-set identiﬁcation, values other than the proposed method are taken from [24].”
Few-Shot Open-Set Recognition Using Meta-Learning (2020, methodology): “We tested both a CNN, denoted as ConvNet, proposed by [22], and ResNet18 [9].”
Classification-Reconstruction Learning for Open-Set Recognition (2018, methodology): “Our backbone network was the same as the one used in [27] that consists of nine convolutional layers and one fully connected layers, except that ours had decoding parts as shown in Eqn.”
Adversarial Reciprocal Points Learning for Open Set Recognition (2021, methodology): “The classifier for this experiment is the same as the neural network used in [38].”