2021 |
Nearest Neighbor Classifier Embedded Network for Active Learning |
Wan et al. |
AAAI |
- |
image classification and object detection |
Confidence ,nearest neighbor classifier , None , Tra , Hard |
CIFAR-10 and CIFAR-100 |
|
2021 |
Agreement-Discrepancy-Selection: Active Learning with Progressive Distribution Alignment |
Fu et al. |
AAAI |
- |
Image Classification |
agreement-discrepancy-selection , CNNs , Advesiral , Tra , Hard |
CIFAR-10 and CIFAR-100 |
|
2021 |
Unsupervised Active Learning via Subspace Learning |
Li et al. |
AAAI |
- |
Classification |
k most representative samples , Matrix Decomposition ,Subspace Learning , Tra , Hard |
HMDB51 (Kuehne et al. 2011) and UCF50, UTKFace (Zhang, Song, and Qi 2017), one med- ical image dataset HAM10000 (Tschandl, Rosendahl, and Kittler 2018), and one wine quality dataset |
|
2021 |
Embodied Visual Active Learning for Semantic Segmentation |
Nilsson et al. |
AAAI |
- |
Semantic Segmentation |
Existing , FCN-inspired deep network , None , Tra , Hard |
Matterport3D |
|
2021 |
An Information‐Theoretic Framework for Unifying Active Learning Problems |
Nguyen et al. |
AAAI |
code |
level set estimation (LSE), Bayesian optimization |
Uncertainty , Gaussian NN , None , Tra , Hard |
Synthetic, one real-world |
|
2021 |
Dialog Policy Learning for Joint Clarification and Active Learning Queries |
Padmakumar and Mooney |
AAAI |
- |
natural language image retrieval, attribute-based clarification |
Uncertainty , MLP , NOne , Tra , Hard |
iMaterialist Fashion Attribute data, create a new dataset |
jointly perform both clarification and active learning in the context of an in- teractive language-based image retrieval task |
2021 |
MTAAL: Multi-Task Adversarial Active Learning for Medical Named Entity Recognition and Normalization |
Zhou et al. |
AAAI |
code |
Medical Named Entity Recognition |
diversity , Adversarial+ Multi-task , Encoder-Decoder , Pre-FT , Hard |
NCBI dataset, BC5CDR |
How- ever, existing models do not take task-specific features for different tasks and diversity of query samples into account. |