Nicholas Moratelli
I am a Research Scientist working on multimodal foundation models, with a focus on vision-language learning, multimodal post-training, and retrieval-augmented reasoning. My research aims to build scalable multimodal systems that are more grounded, factual, and capable of complex image-text understanding.
I hold a Ph.D. in Artificial Intelligence at AImageLab, University of Modena and Reggio Emilia, where I worked on scaling knowledge-grounded vision-language systems, spanning image captioning, visual reasoning, and large multimodal language models.
My work has been published at top-tier venues including CVPR, ACL, ICLR, and BMVC. I previously conducted industrial research at Amazon AGI (Nova), and I am currently building multimodal and video foundation models at Tether.
Latest News
Publications
Benchmarking Deflection and Hallucination in Large Vision-Language Models
In Annual Meeting of the Association for Computational Linguistics, 2026
Augmenting Multimodal LLMs with Self-Reflective Tokens for Knowledge-based Visual Question Answering
In Conference on Computer Vision and Pattern Recognition, 2025
Causal Graphical Models for Vision-Language Compositional Understanding
In International Conference on Learning Representations 2025
Revisiting Image Captioning Training Paradigm via Direct CLIP-based Optimization
In British Machine Vision Conference 2024
Wiki-LLaVA: Hierarchical Retrieval-Augmented Generation for Multimodal LLMs
In Conference on Computer Vision and Pattern Recognition Workshops, 2024
The Revolution of Multimodal Large Language Models: A Survey
In Findings of the Association for Computational Linguistics, 2024
Fluent and Accurate Image Captioning with a Self-Trained Reward Model
In International Conference on Pattern Recognition, 2024
Positive-Augmented Contrastive Learning for Vision-and-Language Evaluation and Training
In International Journal of Computer Vision, 2025
Mitigating Hallucinations in Multimodal LLMs via Object-aware Preference Optimization
In British Machine Vision Conference 2025
LLaVA-MORE: A Comparative Study of LLMs and Visual Backbones for Enhanced Visual Instruction Tuning
In International Conference on Computer Vision, 2025
Personalizing Multimodal Large Language Models for Image Captioning: An Experimental Analysis
In European Conference on Computer Vision and Pattern Recognition, 2024
Fashion-oriented image captioning with external knowledge retrieval and fully attentive gates
In Sensors MDPI, 2023
Are Learnable Prompts the Right Way of Prompting? Adapting Vision-and-Language Models with Memory Optimization
In IEEE Intelligent Systems, 2024
Projects
LLAVA-MORE
A Comparative Study of LLMs and Visual Backbones for Enhanced Visual Instruction Tuning.
Get In Touch
Interested in collaboration?
I'm always open to discussing research ideas, potential collaborations, or opportunities to apply AI in innovative ways.
Email Me