CaMEL: Mean Teacher Learning for Image Captioning


Describing images in natural language is a fundamental step towards the automatic modeling of connections between the visual and textual modalities. In this paper we present CaMEL, a novel Transformer-based architecture for image captioning. Our proposed approach leverages the interaction of two interconnected language models that learn from each other during the training phase. The interplay between the two language models follows a mean teacher learning paradigm with knowledge distillation. Experimentally, we assess the effectiveness of the proposed solution on the COCO dataset and in conjunction with different visual feature extractors. When comparing with existing proposals, we demonstrate that our model provides state-of-the-art caption quality with a significantly reduced number of parameters. According to the CIDEr metric, we obtain a new state of the art on COCO when training without using external data. The source code and trained models are publicly available.

International Conference on Pattern Recognition - ICPR 2022

Model in details:


Qualitative examples:


Full Paper: pdf

Code: github

Please cite with the following BibTeX:

  title={CaMEL: Mean Teacher Learning for Image Captioning},
  author={Manuele, Barraco and Stefanini, Matteo and Cornia, Marcella and Cascianelli, Silvia and Baraldi, Lorenzo and Cucchiara, Rita},
  booktitle={International Conference on Pattern Recognition},
Matteo Stefanini
Matteo Stefanini
PhD Candidate in Artificial Intelligence - TEDx Organizer

I’m a (deep) learner who loves freedom. Working on Deep Learning, Genomics, Gene expression and Vision & Language. Driven to be useful for people.