community
active
leiden_hybrid_concepts
label: haiku
community:leiden_hybrid_concepts-run4-c14-c1Dynamic balancing for multi-task learning
Explores gradient/loss balancing techniques with exponential moving average forgetting rates, evaluated on dense prediction tasks like semantic segmentation.
5 members. Each node is clickable.
Loading graph…
Drawn from 1 source
The papers/notes whose extracted claims & findings make up this cluster.
Bridges (2)
Other communities that share members with this one — cross-cutting threads or papers that sit at the seam between two themes.
Findings (4)
- DB-MTL has similar per-epoch running time to gradient balancing methods on NYUv2, slower than loss balancing methods.Computational efficiency comparison.
- DB-MTL training losses decrease smoothly and gradient norms are lower than EW on NYUv2, indicating training stability.Training stability analysis.
- DB-MTL with EMA forgetting rate β in a wide range performs better than without EMA (β=0) on Office-31.Effect of EMA forgetting rate on performance.
- DB-MTL with SegNet backbone achieves Δp = +8.91 on NYUv2, best among all methods.Performance with a different backbone network.
Claims (1)
- DB-MTL does not affect training stability; losses converge smoothly.Training stability claim.