Publications - data-augmentation

Target-Side Paraphrase Augmentation for Sign Language Translation with Large Language ModelsWorkshop paper

Pedro Dal Bianco, Jean Paul Nunes Reinhold, Oscar Stanchi, Facundo Quiroga, Franco Ronchetti, Ulisses Brisolara Corrêa

GenSign Workshop @ CVPR 2026

A large language model (GPT-4o) writes paraphrases of the target sentences to augment sign language translation data: pre-train on the augmented corpus, then fine-tune on the original references. Evaluated on PHOENIX14T, GSL and LSA-T, with lexical metrics and an LLM-as-a-Judge semantic evaluation.

sign-language-translation data-augmentation llm

Bringing balance to hand shape classification: Mitigating data imbalance through generative modelsJournal article

Gastón Ríos, Pedro Dal Bianco, Franco Ronchetti, Facundo Quiroga, Oscar Stanchi, Santiago Ponte Ahón, Waldo Hasperué

Applied Soft Computing 183, 113586 (Elsevier), 2025

Synthetic images from two GANs, ReACGAN (conditioned on the handshape label) and SPADE (conditioned on pose), used to augment small and unbalanced handshape datasets. Improves the state of the art on the RWTH handshape dataset by 5%.

handshape-recognition data-augmentation generative-models