Jul 14, 2020 · We propose a knowledge distillation based method in this work. We first learn a task-specific model for each task. We then learn the multi-task model.

We propose a knowledge distillation based method in this work. We first learn a task-specific model for each task. We then learn the multi-task model.

We provide code for our method that performs semantic segmentation, depth estimation and surface normal estimation on NYU-v2 dataset using SegNet and MTAN.

In this paper we propose online knowledge distillation for training a multi-task net- work, with similar computation and memory requirements as that of single ...

Allowing separate tasks to converge on their own schedules and using knowledge distillation to maintain performance improves accuracy.

Multi-task learning (MTL) is to learn one single model that performs multiple tasks for achieving good performance on all tasks and lower cost on ...

Jan 5, 2021 · We propose a knowledge distillation based method in this work. We first learn a task-specific model for each task. We then learn the multi-task model.

In this paper, we introduce a new knowledge distillation pro- cedure with an alternative match for MTL of dense predic- tion based on two simple design ...

May 24, 2024 · The naïve approach to partial multi-task learning is sub-optimal due to the lack of all-task annotations for learning joint representations.

We propose an online knowledge distillation method, where single-task networks are trained simultaneously with the MTL network to guide the optimization ...