# Image Difficulty Curriculum for Generative Adversarial Networks (CuGAN)

arXiv: https://arxiv.org/abs/1910.08967 · HTML: https://mlsanity.com/paper/1910.08967/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Oct 2019.

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

- [Curriculum Learning: A Survey](https://mlsanity.com/paper/2101.10382/) (2021)
- [Video Summarization Using Deep Neural Networks: A Survey](https://mlsanity.com/paper/2101.06072/) (2021)
- [Curriculum Direct Preference Optimization for Diffusion and Consistency Models](https://mlsanity.com/paper/2405.13637/) (2024)
- [Learning to Solve Travelling Salesman Problem with Hardness-adaptive Curriculum](https://mlsanity.com/paper/2204.03236/) (2022)
- [Boost Your Human Image Generation Model via Direct Preference Optimization](https://mlsanity.com/paper/2405.20216/) (2024)
- [CyTran: A Cycle-Consistent Transformer with Multi-Level Consistency for Non-Contrast to Contrast CT Translation](https://mlsanity.com/paper/2110.06400/) (2021)
- [Towards Faster Training of Diffusion Models: An Inspiration of A Consistency Phenomenon](https://mlsanity.com/paper/2404.07946/) (2024)
- [Self-paced ensemble learning for speech and audio classification](https://mlsanity.com/paper/2103.11988/) (2021)
- [Imitation Learning from Observation with Automatic Discount Scheduling](https://mlsanity.com/paper/2310.07433/) (2023)
- [Learning to Learn: How to Continuously Teach Humans and Machines](https://mlsanity.com/paper/2211.15470/) (2022)
- [RecRecNet: Rectangling Rectified Wide-Angle Images by Thin-Plate Spline Model and DoF-based Curriculum Learning](https://mlsanity.com/paper/2301.01661/) (2023)
- [Toddler-Guidance Learning: Impacts of Critical Period on Multimodal AI Agents](https://mlsanity.com/paper/2201.04990/) (2022)
- [DiscrimLoss: A Universal Loss for Hard Samples and Incorrect Samples Discrimination](https://mlsanity.com/paper/2208.09884/) (2022)
- [Integrating Curricula with Replays: Its Effects on Continual Learning](https://mlsanity.com/paper/2307.05747/) (2023)
- [Understanding Complexity in VideoQA via Visual Program Generation](https://mlsanity.com/paper/2505.13429/) (2025)
- [A Comprehensive Study of Privacy Risks in Curriculum Learning](https://mlsanity.com/paper/2310.10124/) (2023)
- [Efficient Temporal Consistency in Diffusion-Based Video Editing with Adaptor Modules: A Theoretical Framework](https://mlsanity.com/paper/2504.16016/) (2025)
- [CLearViD: Curriculum Learning for Video Description](https://mlsanity.com/paper/2311.04480/) (2023)
- [ObjBlur: A Curriculum Learning Approach With Progressive Object-Level Blurring for Improved Layout-to-Image Generation](https://mlsanity.com/paper/2404.07564/) (2024)
- [Self-Paced Learning Strategy with Easy Sample Prior Based on Confidence for the Flying Bird Object Detection Model Training](https://mlsanity.com/paper/2412.06306/) (2024)
- [Preview-based Category Contrastive Learning for Knowledge Distillation](https://mlsanity.com/paper/2410.14143/) (2024)
- [Curriculum-DPO++: Direct Preference Optimization via Data and Model Curricula for Text-to-Image Generation](https://mlsanity.com/paper/2602.13055/) (2026)
- [Curriculum-guided Change Detection Training: Toward Accurate Serac Fall Monitoring](https://mlsanity.com/paper/2606.28012/) (2026)

## Abstract (arXiv)

Despite the significant advances in recent years, Generative Adversarial Networks (GANs) are still notoriously hard to train. In this paper, we propose three novel curriculum learning strategies for training GANs. All strategies are first based on ranking the training images by their difficulty scores, which are estimated by a state-of-the-art image difficulty predictor. Our first strategy is to divide images into gradually more difficult batches. Our second strategy introduces a novel curriculum loss function for the discriminator that takes into account the difficulty scores of the real images. Our third strategy is based on sampling from an evolving distribution, which favors the easier images during the initial training stages and gradually converges to a uniform distribution, in which samples are equally likely, regardless of difficulty. We compare our curriculum learning strategies with the classic training procedure on two tasks: image generation and image translation. Our experiments indicate that all strategies provide faster convergence and superior results. For example, our best curriculum learning strategy applied on spectrally normalized GANs (SNGANs) fooled human annotators in thinking that generated CIFAR-like images are real in 25.0% of the presented cases, while the SNGANs trained using the classic procedure fooled the annotators in only 18.4% cases. Similarly, in image translation, the human annotators preferred the images produced by the Cycle-consistent GAN (CycleGAN) trained using curriculum learning in 40.5% cases and those produced by CycleGAN based on classic training in only 19.8% cases, 39.7% cases being labeled as ties.