---
title: "Analysis of Machine Translators on Sentences Generated by"
description: "The authors present this as the first translation-centred evaluation for image captioning. 23 references."
authors: ["Natan Moura", "João Medrado Gondim", "Daniela Barreiro Claro", "Babacar Mane"]
year: 2026
venue: "Proceedings of the 17th International Conference on Computational Processing of Portuguese (PROPOR 2026) - Vol. 1"
source: "https://aclanthology.org/2026.propor-1.36/"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/moura-2026-analysis-machine-translators-sentences-generated/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/moura-2026-analysis-machine-translators-sentences-generated/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/21/) / Analysis of Machine Translators on Sentences Generated by Portuguese Image Captioning Models

# Analysis of Machine Translators on Sentences Generated by Portuguese Image Captioning Models

**Authors:** Natan Moura, João Medrado Gondim, [Daniela Barreiro Claro](https://md.research.elamur.ai/free/en/authors/daniela-barreiro-claro/) and Babacar Mane · **Year:** 2026 · **Venue:** [Proceedings of the 17th International Conference on Computational Processing of Portuguese (PROPOR 2026) - Vol. 1](https://md.research.elamur.ai/free/en/venues/propor/) ([International Conference on Computational Processing of Portuguese 2026](https://md.research.elamur.ai/free/en/venues/propor/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.propor-1.36/) · [PDF](https://aclanthology.org/2026.propor-1.36.pdf)

## Summary
Most image captioning (IC) research targets English, and Portuguese lacks native training resources. This study tests the translate-train strategy, in which an English training corpus is machine-translated and a captioning model is then trained on the result. It asks how the choice of translator affects the captions the model produces. The authors translated Flickr30k into Portuguese three times, using LibreTranslate, Groq (an LLM) and Google Cloud Translation. A translation pipeline with checkpointing produced the translated JSON files. They then trained one AoANet model per translated corpus without changing the architecture.

The evaluation combined several measures:
* monetary cost and translation time;
* automatic metrics on the test set (BLEU, METEOR, CIDEr);
* a human survey comparing captions.

All three translated corpora produced models that generate usable Portuguese captions. They scored below the English baseline on BLEU and CIDEr. Human judges preferred captions from the model trained on Google's translation. This suggests that investing in better translation pays off in caption quality. The authors present this as the first translation-centred evaluation for image captioning.

## Key points
- **Data:** Flickr30k contains 31,783 images and 158,915 captions, with five per image.
- **Translators:**
  - LibreTranslate is OpenNMT-based and self-hosted.
  - Groq used the llama3-8b-8192 model.
  - Google Cloud Translation is a paid NMT service. Translating about 9.7M characters cost roughly 180 USD; the other two tools were free.
- **Translation time:**
  - Google took about 5h 40m.
  - LibreTranslate took about 8h 30m.
  - Groq took about 24h, spread over three days because of daily request limits.
- **Human survey:** 88 responses covered 30 images.
  - In 56.7% of images (17), at least one caption was judged to describe the image.
  - Translator preference scores were Google 16, Groq 9 and LibreTranslate 7. On the conforming subset they were 10, 6 and 3.
- **Limitations:**
  - Models were not trained to loss convergence because of cost.
  - Hyperparameters and the Stanford PTB Tokenizer were kept from the English setup.
  - Evaluators were all Brazilian, which may have penalised LibreTranslate's European Portuguese output.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (2): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works

🧭 Topics (1): [🖼️ Vision & language](https://md.research.elamur.ai/free/en/topics/multimodal/) · 657 works

## Cite

Cite as: Natan Moura, João Medrado Gondim, Daniela Barreiro Claro and Babacar Mane. 2026. Analysis of Machine Translators on Sentences Generated by Portuguese Image Captioning Models. Proceedings of the 17th International Conference on Computational Processing of Portuguese (PROPOR 2026) - Vol. 1.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{moura2026analysis,
  title = {{Analysis of Machine Translators on Sentences Generated by Portuguese Image Captioning Models}},
  author = {Natan Moura, João Medrado Gondim, Daniela Barreiro Claro and Babacar Mane},
  booktitle = {Proceedings of the 17th International Conference on Computational Processing of Portuguese (PROPOR 2026) - Vol. 1},
  year = {2026},
  url = {https://aclanthology.org/2026.propor-1.36/}
}
```

- [← previous Analysis of Child-Caregiver Interactions for Developing a Caregiver Spoken Dialogue System](https://md.research.elamur.ai/free/en/papers/yamashita-2026-analysis-child-caregiver-interactions-developing/)
- [next → Analysis of the Neglect-Zero Effect in Large Language Models](https://md.research.elamur.ai/free/en/papers/tanaka-2026-analysis-neglect-zero-effect-large/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/21/)

## References

23 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. This work was supported by FAPESB (TIC002/2015 and CCE022/2023), CAPES, CNPQ, and INCT-TILDIAR/CNPq (408490/2024-1). [find](https://scholar.google.com/scholar?q=This+work+was+supported+by+FAPESB+%28TIC002%2F2015+and+CCE022%2F2023%29%2C+CAPES%2C+CNPQ%2C+and+INCT%2DTILDIAR%2FCNPq+%28408490%2F2024%2D1%29%2E)
2. Raffaella Bernardi, Ruket Cakici, Desmond Elliott, Aykut Erdem, Erkut Erdem, Nazli Ikizler-Cinbis, Frank Keller, Adrian Muscat, and Barbara Plank. 2016. Automatic description generation from images: A survey of models, datasets, and evaluation measures. Journal of Artificial Intelligence Research, 5… [find](https://scholar.google.com/scholar?q=Raffaella+Bernardi%2C+Ruket+Cakici%2C+Desmond+Elliott%2C+Aykut+Erdem%2C+Erkut+Erdem%2C+Nazli+Ikizler%2DCinbis%2C+Frank+Keller%2C+Adrian+Muscat%2C+and+Barbara+Plank%2E+2016%2E+Automatic+description+generation+from+images%3A+A)
3. Gabriel Oliveira dos Santos, Diego Alysson Braga Moreira, Alef Iury Ferreira, Jhessica Silva, Luiz Pereira, Pedro Bueno, Thiago Sousa, Helena Maia, Nádia Da Silva, Esther Colombini, Helio Pedrini, and Sandra Avila. 2023. CAPIVARA: Cost-efficient approach for improving multilingual CLIP performance o… [find](https://scholar.google.com/scholar?q=Gabriel+Oliveira+dos+Santos%2C+Diego+Alysson+Braga+Moreira%2C+Alef+Iury+Ferreira%2C+Jhessica+Silva%2C+Luiz+Pereira%2C+Pedro+Bueno%2C+Thiago+Sousa%2C+Helena+Maia%2C+N%C3%A1dia+Da+Silva%2C+Esther+Colombini%2C+Helio+Pedrini%2C+and)
4. Amir El-Komy, Osama R Shahin, Rasha M Abd El-Aziz, and Ahmed I Taloba. 2022. Integration of computer vision and natural language processing in multimedia robotics application. Inf. Sci, 7(6):1–12. [find](https://scholar.google.com/scholar?q=Amir+El%2DKomy%2C+Osama+R+Shahin%2C+Rasha+M+Abd+El%2DAziz%2C+and+Ahmed+I+Taloba%2E+2022%2E+Integration+of+computer+vision+and+natural+language+processing+in+multimedia+robotics+application%2E+Inf%2E+Sci%2C+7%286%29%3A1%E2%80%9312%2E)
5. Albert Gatt and Emiel Krahmer. 2018. Survey of the state of the art in natural language generation: Core tasks, applications and evaluation. Journal of Artificial Intelligence Research, 61:65–170. [find](https://scholar.google.com/scholar?q=Albert+Gatt+and+Emiel+Krahmer%2E+2018%2E+Survey+of+the+state+of+the+art+in+natural+language+generation%3A+Core+tasks%2C+applications+and+evaluation%2E+Journal+of+Artificial+Intelligence+Research%2C+61%3A65%E2%80%93170%2E)
6. João Gondim, Daniela Claro, and Marlo Souza. 2022. Towards image captioning for the portuguese language: Evaluation on a translated dataset. In Proceedings of the 24th International Conference on Enterprise Information Systems - Volume 2: ICEIS, pages 384–393. INSTICC, SciTePress. [find](https://scholar.google.com/scholar?q=Jo%C3%A3o+Gondim%2C+Daniela+Claro%2C+and+Marlo+Souza%2E+2022%2E+Towards+image+captioning+for+the+portuguese+language%3A+Evaluation+on+a+translated+dataset%2E+In+Proceedings+of+the+24th+International+Conference+on+Ente)
7. João Gondim, Daniela Barreiro Claro, and Marlo Souza. 2025. A bilingual analysis of multi-head attention mechanism for image captioning based on morphosyntactic information. Journal of the Brazilian Computer Society, 31(1):1063–1076. [find](https://scholar.google.com/scholar?q=Jo%C3%A3o+Gondim%2C+Daniela+Barreiro+Claro%2C+and+Marlo+Souza%2E+2025%2E+A+bilingual+analysis+of+multi%2Dhead+attention+mechanism+for+image+captioning+based+on+morphosyntactic+information%2E+Journal+of+the+Brazilian+C)
8. Google. 2008. Cloud translation. Accessed on August 02, 2024. [find](https://scholar.google.com/scholar?q=Google%2E+2008%2E+Cloud+translation%2E+Accessed+on+August+02%2C+2024%2E)
9. Groq. 2016. Groq. Accessed on August 02, 2024. [find](https://scholar.google.com/scholar?q=Groq%2E+2016%2E+Groq%2E+Accessed+on+August+02%2C+2024%2E)
10. Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. [Deep residual learning for image recognition](https://md.research.elamur.ai/free/en/papers/he-2015-deep-residual-learning-image-recognition/). In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778.
11. Lun Huang, Wenmin Wang, Jie Chen, and Xiao-Yong Wei. 2019. Attention on attention for image captioning. In 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pages 4633–4642. [find](https://scholar.google.com/scholar?q=Lun+Huang%2C+Wenmin+Wang%2C+Jie+Chen%2C+and+Xiao%2DYong+Wei%2E+2019%2E+Attention+on+attention+for+image+captioning%2E+In+2019+IEEE%2FCVF+International+Conference+on+Computer+Vision+%28ICCV%29%2C+pages+4633%E2%80%934642%2E)
12. Guillaume Klein, Yoon Kim, Yuntian Deng, Jean Senellart, and Alexander Rush. 2017. [OpenNMT: Open-source toolkit for neural machine translation](https://md.research.elamur.ai/free/en/papers/klein-2017-opennmt-open-source-toolkit-neural/). In Proceedings of ACL 2017, System Demonstrations, pages 67–72, Vancouver, Canada. Association for Computational Linguistics.
13. Peter Young, Alice Lai, Micah Hodosh, and Julia Hockenmaier. 2014. From image descriptions to visual denotations: New similarity metrics for semantic inference over event descriptions. TACL, 2:67–78. [find](https://scholar.google.com/scholar?q=Peter+Young%2C+Alice+Lai%2C+Micah+Hodosh%2C+and+Julia+Hockenmaier%2E+2014%2E+From+image+descriptions+to+visual+denotations%3A+New+similarity+metrics+for+semantic+inference+over+event+descriptions%2E+TACL%2C+2%3A67%E2%80%9378%2E)
14. Alon Lavie and Abhaya Agarwal. 2007. [Meteor: An automatic metric for mt evaluation with high levels of correlation with human judgments](https://md.research.elamur.ai/free/en/papers/lavie-2007-meteor-automatic-metric-mt-evaluation/). In Proceedings of the Second Workshop on Statistical Machine Translation, StatMT ’07, page 228–231, USA. Association for Computational Linguistics.
15. Elizabeth D. Liddy. 2001. Natural language processing. In Encyclopedia of Library and Information Science, 2nd edition. Marcel Decker, Inc., NY. [find](https://scholar.google.com/scholar?q=Elizabeth+D%2E+Liddy%2E+2001%2E+Natural+language+processing%2E+In+Encyclopedia+of+Library+and+Information+Science%2C+2nd+edition%2E+Marcel+Decker%2C+Inc%2E%2C+NY%2E)
16. Tsung-Yi Lin, Michael Maire, Serge J. Belongie, Lubomir D. Bourdev, Ross B. Girshick, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C. Lawrence Zitnick. 2014. [Microsoft COCO: common objects in context](https://md.research.elamur.ai/free/en/papers/lin-2014-microsoft-coco-common-objects-context/). CoRR, abs/1405.0312.
17. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. [Bleu: A method for automatic evaluation of machine translation](https://md.research.elamur.ai/free/en/papers/papineni-2002-bleu-method-automatic-evaluation-machine/). In Proceedings of the 40th Annual Meeting on Association for Computational Linguistics, ACL ’02, page 311–318, USA. Association for Computational Linguistics.
18. Guilherme Moraes Rosa, Luiz Henrique Bonifacio, Leandro Rodrigues de Souza, Roberto Lotufo, and Rodrigo Nogueira. 2021. A cost-benefit analysis of cross-lingual transfer methods. In Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, pages 4231–4242, Online. Association for C… [find](https://scholar.google.com/scholar?q=Guilherme+Moraes+Rosa%2C+Luiz+Henrique+Bonifacio%2C+Leandro+Rodrigues+de+Souza%2C+Roberto+Lotufo%2C+and+Rodrigo+Nogueira%2E+2021%2E+A+cost%2Dbenefit+analysis+of+cross%2Dlingual+transfer+methods%2E+In+Findings+of+the+As)
19. Gabriel Oliveira dos Santos, Esther Luna Colombini, and Sandra Avila. 2021. Pracegover: A large dataset for image captioning in portuguese. In Proceedings of the 34th Conference on Graphics, Patterns and Images (SIBGRAPI). IEEE. [find](https://scholar.google.com/scholar?q=Gabriel+Oliveira+dos+Santos%2C+Esther+Luna+Colombini%2C+and+Sandra+Avila%2E+2021%2E+Pracegover%3A+A+large+dataset+for+image+captioning+in+portuguese%2E+In+Proceedings+of+the+34th+Conference+on+Graphics%2C+Patterns+)
20. Richard Szeliski. 2010. Computer Vision: Algorithms and Applications, 1st edition. Springer-Verlag, Berlin, Heidelberg. [find](https://scholar.google.com/scholar?q=Richard+Szeliski%2E+2010%2E+Computer+Vision%3A+Algorithms+and+Applications%2C+1st+edition%2E+Springer%2DVerlag%2C+Berlin%2C+Heidelberg%2E)
21. Ramakrishna Vedantam, C. Lawrence Zitnick, and Devi Parikh. 2014. [Cider: Consensus-based image description evaluation](https://md.research.elamur.ai/free/en/papers/vedantam-2014-cider-consensus-based-image-description/). CoRR, abs/1411.5726.
22. V H Vishnu Kumar and N Lalithamani. 2022. English to tamil multi-modal image captioning translation. In 2022 IEEE World Conference on Applied Intelligence and Computing (AIC), pages 332–338. [find](https://scholar.google.com/scholar?q=V+H+Vishnu+Kumar+and+N+Lalithamani%2E+2022%2E+English+to+tamil+multi%2Dmodal+image+captioning+translation%2E+In+2022+IEEE+World+Conference+on+Applied+Intelligence+and+Computing+%28AIC%29%2C+pages+332%E2%80%93338%2E)
23. Yonghui Wu, Mike Schuster, Zhifeng Chen, Quoc V. Le, Mohammad Norouzi, Wolfgang Macherey, Maxim Krikun, Yuan Cao, Qin Gao, Klaus Macherey, Jeff Klingner, Apurva Shah, Melvin Johnson, Xiaobing Liu, Łukasz Kaiser, Stephan Gouws, Yoshikiyo Kato, Taku Kudo, Hideto Kazawa, and 12 others. 2016. Google’s n… [find](https://scholar.google.com/scholar?q=Yonghui+Wu%2C+Mike+Schuster%2C+Zhifeng+Chen%2C+Quoc+V%2E+Le%2C+Mohammad+Norouzi%2C+Wolfgang+Macherey%2C+Maxim+Krikun%2C+Yuan+Cao%2C+Qin+Gao%2C+Klaus+Macherey%2C+Jeff+Klingner%2C+Apurva+Shah%2C+Melvin+Johnson%2C+Xiaobing+Liu%2C+%C5%81uk)

*Updated 1 November 2026*