| Year | Title | Author | Publication | Code | Tasks | Notes | Datasets | Notions |
|---|---|---|---|---|---|---|---|---|
| 2022 | Active Learning for Open-set Annotation | Ning et al. | CVPR | - | Image Classification | Uncertainty, DNNs, None, Tra, Filter+Annotate |
CIFAR10, CIFAR100, Tiny-Imagenet | The goal is to precisely filter out examples from unknown classes, while actively select a query set that contains examples from known classes as pure as possi- ble. |
| 2022 | Towards Robust and Reproducible Active Learning using Neural Networks | Munjal et al. | CVPR | code | A framework to solve unstable performance | Any |
||
| 2022 | Towards Fewer Annotations: Active Learning via Region Impurity and Prediction Uncertainty for Domain Adaptive Semantic Segmentation | Xie et al. | CVPR | code | Semantic Segmentation | Uncertainty,DeepLab-v2, DeepLab-v3+, Domain Adaptation, PT+FT, Soft |
GTAV, SYNTHIA | |
| 2022 | Entropy-based Active Learning for Object Detection with Progressive Diversity Constraint | Wu et al. | CVPR | - | Object detection | Entropy+Diversity, ResNet-50, None, FT+PT, Hard |
MS COCO, Pascal VOC | |
| 2022 | Meta Agent Teaming Active Learning for Pose Estimation | Gong et al. | CVPR | - | Pose Estimation | Informative+diverse, Markov Decision Process, Reinforce learning, Tra, Hard |
ICVL [46], NYU [48] and BigHand2.2M | |
| 2022 | Active Learning by Feature Mixing | Parvaneh et al. | CVPR | code | CLassification | informative, Pre-trained model , Representation-level, PT+FT, None |
MNIST [23], EM- NIST [9], CIFAR10 [21], CIFAR100 [21], Mini- ImageNet [32], DomianNet-Real [30] as well as two subsets of DomainNet-Real for image datasets. |