Understanding Episode Hardness in Few-Shot Learning
Achieving generalization for deep learning models has usually suffered from the bottleneck of annotated sample scarcity. As a common way of tackling this issue, few-shot learning focuses on "episodes", i.e. sampled tasks that help the model acquire generalizable knowledge onto unseen categ...
Ausführliche Beschreibung
Bibliographische Detailangaben
| Veröffentlicht in: | IEEE transactions on pattern analysis and machine intelligence. - 1979. - PP(2024) vom: 08. Okt.
|
|---|
| 1. Verfasser: |
Guo, Yurong
(VerfasserIn) |
|---|
| Weitere Verfasser: |
Du, Ruoyi,
Sain, Aneeshan,
Liang, Kongming,
Dong, Yuan,
Song, Yi-Zhe,
Ma, Zhanyu |
|---|
| Format: | Online-Aufsatz
|
|---|
| Sprache: | English |
|---|
| Veröffentlicht: |
2024
|
|---|
| Zugriff auf das übergeordnete Werk: | IEEE transactions on pattern analysis and machine intelligence
|
|---|
| Schlagworte: | Journal Article |
|---|