# Regularizing CNN Transfer Learning with Randomised Regression

arXiv: https://arxiv.org/abs/1908.05997 · HTML: https://mlsanity.com/paper/1908.05997/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 Aug 2019.

## Built on by (most cited first, 5 of 5)

- [Post-pre-training for Modality Alignment in Vision-Language Foundation Models](https://mlsanity.com/paper/2504.12717/) (2025)
- [ODMTCNet: An Interpretable Multi-view Deep Neural Network Architecture for Image Feature Representation](https://mlsanity.com/paper/2110.14830/) (2021)
- [Adaptive Random Feature Regularization on Fine-tuning Deep Neural Networks](https://mlsanity.com/paper/2403.10097/) (2024)
- [One to Transfer All: A Universal Transfer Framework for Vision Foundation Model with Few Data](https://mlsanity.com/paper/2111.12386/) (2021)
- [Black-Box Continual Learning for Vision-Language Models](https://mlsanity.com/paper/2606.22999/) (2026)

## Abstract (arXiv)

This paper is about regularizing deep convolutional networks (CNNs) based on an adaptive framework for transfer learning with limited training data in the target domain. Recent advances of CNN regularization in this context are commonly due to the use of additional regularization objectives. They guide the training away from the target task using some forms of concrete tasks. Unlike those related approaches, we suggest that an objective without a concrete goal can still serve well as a regularized. In particular, we demonstrate Pseudo-task Regularization (PtR) which dynamically regularizes a network by simply attempting to regress image representations to pseudo-regression targets during fine-tuning. That is, a CNN is efficiently regularized without additional resources of data or prior domain expertise. In sum, the proposed PtR provides: a) an alternative for network regularization without dependence on the design of concrete regularization objectives or extra annotations; b) a dynamically adjusted and maintained strength of regularization effect by balancing the gradient norms between objectives on-line. Through numerous experiments, surprisingly, the improvements on classification accuracy by PtR are shown greater or on a par to the recent state-of-the-art methods.