A new Romanian Flickr30k translation plus synthetic VQA corpus, and LoRA-fine-tuned VLMs that improve Romanian VQA and captioning for LLaMA-3.2 and Qwen2-VL, but not LLaVA-1.6.
X-LLaV A: Optimizing bilingual large vision-language alignment
1 Pith paper cite this work, alongside 2 external citations. Polarity classification is still indexing.
1
Pith paper citing it
2
external citations · OpenAlex
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Parameter Efficient Multimodal Instruction Tuning for Romanian Vision Language Models
A new Romanian Flickr30k translation plus synthetic VQA corpus, and LoRA-fine-tuned VLMs that improve Romanian VQA and captioning for LLaMA-3.2 and Qwen2-VL, but not LLaVA-1.6.