Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
157 tokens/sec
GPT-4o
43 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Developmental Pretraining (DPT) for Image Classification Networks (2312.00304v1)

Published 1 Dec 2023 in cs.LG and cs.CV

Abstract: In the backdrop of increasing data requirements of Deep Neural Networks for object recognition that is growing more untenable by the day, we present Developmental PreTraining (DPT) as a possible solution. DPT is designed as a curriculum-based pre-training approach designed to rival traditional pre-training techniques that are data-hungry. These training approaches also introduce unnecessary features that could be misleading when the network is employed in a downstream classification task where the data is sufficiently different from the pre-training data and is scarce. We design the curriculum for DPT by drawing inspiration from human infant visual development. DPT employs a phased approach where carefully-selected primitive and universal features like edges and shapes are taught to the network participating in our pre-training regime. A model that underwent the DPT regime is tested against models with randomised weights to evaluate the viability of DPT.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (22)
  1. Imagenet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition, pages 248–255, 2009.
  2. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016.
  3. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014.
  4. Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning, pages 6105–6114. PMLR, 2019.
  5. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929, 2020.
  6. Review of deep learning: Concepts, cnn architectures, challenges, applications, future directions. Journal of big Data, 8:1–74, 2021.
  7. A study of cnn and transfer learning in medical imaging: Advantages, challenges, future scope. Sustainability, 15(7):5930, 2023.
  8. A comprehensive survey on transfer learning. Proceedings of the IEEE, 109(1):43–76, 2021.
  9. Curriculum learning. In Proceedings of the 26th annual international conference on machine learning, pages 41–48, 2009.
  10. A survey on curriculum learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(9):4555–4576, 2021.
  11. Baby steps: How “less is more” in unsupervised dependency parsing. 2009.
  12. Self-paced curriculum learning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 29, 2015.
  13. Lessons from infant learning for unsupervised machine learning. Nature Machine Intelligence, 4(6):510–520, 2022.
  14. A developmental approach to machine learning? Frontiers in psychology, 8:2124, 2017.
  15. Recurrent neural circuits for contour detection. arXiv preprint arXiv:2010.15314, 2020.
  16. Potential downside of high initial visual acuity. Proceedings of the National Academy of Sciences, 115(44):11333–11338, 2018.
  17. Dense extreme inception network for edge detection. Pattern Recognition, 139:109461, 2023.
  18. A El Korchi. 2d geometric shapes dataset. Mendeley Data, 1:2020, 2020.
  19. Jeremy Howard. imagenette.
  20. Dropout: a simple way to prevent neural networks from overfitting. The journal of machine learning research, 15(1):1929–1958, 2014.
  21. Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms. arXiv preprint arXiv:1708.07747, 2017.
  22. Vggface2: A dataset for recognising faces across pose and age. In 2018 13th IEEE international conference on automatic face & gesture recognition (FG 2018), pages 67–74. IEEE, 2018.

Summary

We haven't generated a summary for this paper yet.