2 citations · 4 across the 8 of their papers we have counts for
10 papers
Question-Guided Evidence Acquisition for Multimodal Visual Question Answering
Alin-Ionut Popa
Multimodal LLMs can see a document, but they often can't read it reliably. Small text, tables, visual cues, and topological elements still trip them up under direct visual inferenc…
PIVOT: Bridging Planning and Execution in LLM Agents via Trajectory Refinement
Tuo Zhang, Alin-Ionut Popa, Yan Xu +2
Large language model (LLM)-based agents frequently generate seemingly coherent plans that fail upon execution due to infeasible actions, constraint violations, and compounding erro…
Bidirectional Long-Range Parser for Sequential Data Understanding
George Leotescu, Daniel Voinea, Alin-Ionut Popa
The transformer is a powerful data modelling framework responsible for remarkable performance on a wide range of tasks. However, they are limited in terms of scalability as it is s…
Watermark Text Pattern Spotting in Document Images
Mateusz Krubiński, Stefan Matcovici, Diana Grigore +2
Watermark text spotting in document images can offer access to an often unexplored source of information, providing crucial evidence about a record's scope, audience and sometimes…
Reconstructing Three-Dimensional Models of Interacting Humans
Mihai Fieraru, Mihai Zanfir, Elisabeta Oneata +3
Understanding 3d human interactions is fundamental for fine-grained scene analysis and behavioural modeling. However, most of the existing models predict incorrect, lifeless 3d est…
CONSENT: Context Sensitive Transformer for Bold Words Classification
Ionut-Catalin Sandu, Daniel Voinea, Alin-Ionut Popa
We present CONSENT, a simple yet effective CONtext SENsitive Transformer framework for context-dependent object classification within a fully-trainable end-to-end deep learning pip…