OSDTW decomposes long-tailed recognition into head and tail tasks with shared encoder and derives a Fisher-based bias-variance proxy to select shared depth and task weights.
Advances in Neural Information Pro- cessing Systems32(2019)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
OSDTW: Optimal Shared Depth and Task Weighting for Long-Tailed Recognition
OSDTW decomposes long-tailed recognition into head and tail tasks with shared encoder and derives a Fisher-based bias-variance proxy to select shared depth and task weights.