# Meta-Transfer Learning through Hard Tasks

arXiv: https://arxiv.org/abs/1910.03648 · HTML: https://mlsanity.com/paper/1910.03648/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Oct 2019.

## Built on by (most cited first, 23 of 23)

- [Few-Shot Incremental Learning with Continually Evolved Classifiers](https://mlsanity.com/paper/2104.03047/) (2021)
- [Overcoming Catastrophic Forgetting in Incremental Few-Shot Learning by Finding Flat Minima](https://mlsanity.com/paper/2111.01549/) (2021)
- [How to Train Your MAML to Excel in Few-Shot Classification](https://mlsanity.com/paper/2106.16245/) (2021)
- [LibFewShot: A Comprehensive Library for Few-shot Learning](https://mlsanity.com/paper/2109.04898/) (2021)
- [Exploring Hierarchical Graph Representation for Large-Scale Zero-Shot Image Classification](https://mlsanity.com/paper/2203.01386/) (2022)
- [Reliable Few-shot Learning under Dual Noises](https://mlsanity.com/paper/2506.16330/) (2025)
- [Prototype Completion for Few-Shot Learning](https://mlsanity.com/paper/2108.05010/) (2021)
- [Uniform Sampling over Episode Difficulty](https://mlsanity.com/paper/2108.01662/) (2021)
- [Sharing to learn and learning to share; Fitting together Meta-Learning, Multi-Task Learning, and Transfer Learning: A meta review](https://mlsanity.com/paper/2111.12146/) (2021)
- [Contextualizing Meta-Learning via Learning to Decompose](https://mlsanity.com/paper/2106.08112/) (2021)
- [KTN: Knowledge Transfer Network for Learning Multi-person 2D-3D Correspondences](https://mlsanity.com/paper/2206.10090/) (2022)
- [DAC-MR: Data Augmentation Consistency Based Meta-Regularization for Meta-Learning](https://mlsanity.com/paper/2305.07892/) (2023)
- [Understanding Knowledge Transferability for Transfer Learning: A Survey](https://mlsanity.com/paper/2507.03175/) (2025)
- [A Survey of Deep Visual Cross-Domain Few-Shot Learning](https://mlsanity.com/paper/2303.09253/) (2023)
- [Task Attended Meta-Learning for Few-Shot Learning](https://mlsanity.com/paper/2106.10642/) (2021)
- [Bayesian Evidential Learning for Few-Shot Classification](https://mlsanity.com/paper/2207.13137/) (2022)
- [Rethinking the Number of Shots in Robust Model-Agnostic Meta-Learning](https://mlsanity.com/paper/2211.15180/) (2022)
- [Task Attribute Distance for Few-Shot Learning: Theoretical Analysis and Applications](https://mlsanity.com/paper/2403.03535/) (2024)
- [A Study on Representation Transfer for Few-Shot Learning](https://mlsanity.com/paper/2209.02073/) (2022)
- [Improving Generalization in Meta-Learning via Meta-Gradient Augmentation](https://mlsanity.com/paper/2306.08460/) (2023)
- [Learn To Learn More Precisely](https://mlsanity.com/paper/2408.04590/) (2024)
- [Oops, I Sampled it Again: Reinterpreting Confidence Intervals in Few-Shot Learning](https://mlsanity.com/paper/2409.02850/) (2024)
- [Few to Big: Prototype Expansion Network via Diffusion Learner for Point Cloud Few-shot Semantic Segmentation](https://mlsanity.com/paper/2509.12878/) (2025)

## Abstract (arXiv)

Meta-learning has been proposed as a framework to address the challenging few-shot learning setting. The key idea is to leverage a large number of similar few-shot tasks in order to learn how to adapt a base-learner to a new task for which only a few labeled samples are available. As deep neural networks (DNNs) tend to overfit using a few samples only, typical meta-learning models use shallow neural networks, thus limiting its effectiveness. In order to achieve top performance, some recent works tried to use the DNNs pre-trained on large-scale datasets but mostly in straight-forward manners, e.g., (1) taking their weights as a warm start of meta-training, and (2) freezing their convolutional layers as the feature extractor of base-learners. In this paper, we propose a novel approach called meta-transfer learning (MTL) which learns to transfer the weights of a deep NN for few-shot learning tasks. Specifically, meta refers to training multiple tasks, and transfer is achieved by learning scaling and shifting functions of DNN weights for each task. In addition, we introduce the hard task (HT) meta-batch scheme as an effective learning curriculum that further boosts the learning efficiency of MTL. We conduct few-shot learning experiments and report top performance for five-class few-shot recognition tasks on three challenging benchmarks: miniImageNet, tieredImageNet and Fewshot-CIFAR100 (FC100). Extensive comparisons to related works validate that our MTL approach trained with the proposed HT meta-batch scheme achieves top performance. An ablation study also shows that both components contribute to fast convergence and high accuracy.