arXiv subjects

João Paulo Papa

Publications and source records attributed to João Paulo Papa.

At least 19 recordsLinked to original sources

arXiv2026

Audit Before You Distill: A Controlled Workflow for Reasoning Transfer into Compact Medical Language Models

Teacher rationales can help a smaller model answer health professions questions, but they can also reinforce errors. We examined when they help and how this evidence can guide training using Brazilian Portuguese questions from medical and multiprofessional residency examinations. First, we measured how a teacher rationale changed the student's probability of selecting the correct answer and compared this effect with rationales drawn from other questions. The complete rationale increased this probability by 21~p.p. when the teacher was correct and decreased it by 14~p.p. when the teacher was wrong. By intervening on the reasoning content, we showed that shuffled rationales did not produce the same pattern. Guided by this result, we trained only with rationales attached to correct teacher answers and compared answer-only training, step-by-step distillation, and our proposed Distill-SFT method. Distill-SFT showed a modest advantage over step-by-step distillation on previously missed questions under reasoning, while using fewer training instances by combining the answer and rationale into a single target. We also investigated catastrophic forgetting using the out-of-domain ENEM dataset. Answer-only training reduced reasoning accuracy by 12~p.p. relative to baseline, compared with approximately 4~p.p. for either rationale-based training method. Together, these findings support a workflow that audits rationale content before distillation, verifies teacher answers before supervision, and trains compact models efficiently for health professions question answering.

cs.CL↗

arXiv2026

SPACE: Semantic Projection and Alignment of CLIP Embeddings for Domain Adaptation

A fundamental challenge in deploying vision models is domain shift, which arises when training and test data follow different distributions, leading to degraded performance. This challenge is amplified when the same semantic concept appears under distinct visual forms, such as photographs and sketches, where visual similarity is weak despite semantic correspondence. Existing unsupervised domain-adaptation methods aim to align distributions across domains but often ignore semantic relationships among samples of the same class. To address this issue, this paper introduces SPACE, a method that exploits the semantic structure of CLIP's vision-language space for domain adaptation. The key idea is to use text descriptions as semantic anchors by applying Singular Value Decomposition to CLIP embeddings of class descriptions, yielding an orthogonal basis that captures semantic relationships among categories. Visual features from both domains are projected into this semantic subspace, aligning images based on meaning rather than appearance.

cs.CV↗

arXiv2026

On the Complementarity of Quantum and Classical Features: Adaptive Hybrid Quantum-Classical Feature Fusion for Breast Cancer Classification

The integration of quantum machine learning with classical deep learning offers promising avenues for medical image analysis by mapping data into high-dimensional Hilbert spaces. However, effectively unifying these distinct paradigms remains challenging due to common optimization asymmetries. In this paper, a novel hybrid quantum-classical architecture for breast cancer diagnosis based on a dual-branch feature-extraction pipeline is proposed. Our framework extracts and unifies complementary representations from classical models and quantum circuits, exploring both trainable and deterministic (non-trainable) quantum paradigms. To integrate these embeddings, three progressive feature fusion strategies are introduced: Static Hybrid Fusion (SHF) for offline extraction, Dynamic Hybrid Fusion (DHF) for end-to-end co-adaptation, and a novel Temperature-Scaled Hybrid Fusion (TSHF). The TSHF strategy incorporates a learnable scalar, inspired by multimodal learning, that dynamically balances hybrid gradient dynamics and resolves optimization bottlenecks. Empirical validation on the BreastMNIST dataset confirms our hypothesis that unifying diverse feature representations creates a richer data context. The TSHF strategy, specifically when pairing a ResNet backbone with a trainable quantum circuit, achieved a peak accuracy of 87.82%, F1-score of 91.77%, and an AUC-ROC of 89.08%, outperforming purely classical baselines. These results demonstrate that the proposed hybrid framework improves classification accuracy and threshold reliability, providing a stable, high-performance architecture for the clinical deployment of quantum-enhanced diagnostic tools.

cs.CV↗

arXiv2026

Comparative validation of surgical phase recognition, instrument keypoint estimation, and instrument instance segmentation in endoscopy: Results of the PhaKIR 2024 challenge

Reliable recognition and localization of surgical instruments in endoscopic video recordings are foundational for a wide range of applications in computer- and robot-assisted minimally invasive surgery (RAMIS), including surgical training, skill assessment, and autonomous assistance. However, robust performance under real-world conditions remains a significant challenge. Incorporating surgical context - such as the current procedural phase - has emerged as a promising strategy to improve robustness and interpretability. To address these challenges, we organized the Surgical Procedure Phase, Keypoint, and Instrument Recognition (PhaKIR) sub-challenge as part of the Endoscopic Vision (EndoVis) challenge at MICCAI 2024. We introduced a novel, multi-center dataset comprising thirteen full-length laparoscopic cholecystectomy videos collected from three distinct medical institutions, with unified annotations for three interrelated tasks: surgical phase recognition, instrument keypoint estimation, and instrument instance segmentation. Unlike existing datasets, ours enables joint investigation of instrument localization and procedural context within the same data while supporting the integration of temporal information across entire procedures. We report results and findings in accordance with the BIAS guidelines for biomedical image analysis challenges. The PhaKIR sub-challenge advances the field by providing a unique benchmark for developing temporally aware, context-driven methods in RAMIS and offers a high-quality resource to support future research in surgical scene understanding.

cs.CV↗

arXiv2025

Warm-Starting QAOA with XY Mixers: A Novel Approach for Quantum-Enhanced Vehicle Routing Optimization

Quantum optimization algorithms, such as the Quantum Approximate Optimization Algorithm, are emerging as promising heuristics for solving complex combinatorial problems. To improve performance, several extensions to the standard QAOA framework have been proposed in recent years. Two notable directions include: warm-starting techniques, which incorporate classical approximate solutions to guide the quantum evolution, and custom mixer Hamiltonians, such as XY mixers, which constrain the search to feasible subspaces aligned with the structure of the problem. In this work, we propose an approach that integrates these two strategies: a warm-start initialization with an XY mixer ansatz, enabling constraint-preserving quantum evolution biased toward high-quality classical solutions. The method begins by reformulating the combinatorial problem as a MaxCut instance, solved approximately using the Goemans-Williamson algorithm. The resulting binary solution is relaxed and used to construct a biased superposition over valid one-hot quantum states, maintaining compatibility with the XY mixer's constraints. We evaluate the approach on 5-city instances of the Traveling Salesperson Problem, a canonical optimization problem frequently encountered as a subroutine in real-world Vehicle Routing Problems. Our method is benchmarked against both the standard XY-mixer QAOA and a warm-start-only variant based on MaxCut relaxation. Results show that the proposed combination consistently outperforms both baselines in terms of the percentage and rank of optimal solutions, demonstrating the effectiveness of combining structured initializations with constraint-aware quantum evolution for optimization problems.

quant-ph↗

arXiv2024

A Review on Scientific Knowledge Extraction using Large Language Models in Biomedical Sciences

The rapid advancement of large language models (LLMs) has opened new boundaries in the extraction and synthesis of medical knowledge, particularly within evidence synthesis. This paper reviews the state-of-the-art applications of LLMs in the biomedical domain, exploring their effectiveness in automating complex tasks such as evidence synthesis and data extraction from a biomedical corpus of documents. While LLMs demonstrate remarkable potential, significant challenges remain, including issues related to hallucinations, contextual understanding, and the ability to generalize across diverse medical tasks. We highlight critical gaps in the current research literature, particularly the need for unified benchmarks to standardize evaluations and ensure reliability in real-world applications. In addition, we propose directions for future research, emphasizing the integration of state-of-the-art techniques such as retrieval-augmented generation (RAG) to enhance LLM performance in evidence synthesis. By addressing these challenges and utilizing the strengths of LLMs, we aim to improve access to medical literature and facilitate meaningful discoveries in healthcare.

cs.CL↗

arXiv2024

BioNeRF: Biologically Plausible Neural Radiance Fields for View Synthesis

This paper presents BioNeRF, a biologically plausible architecture that models scenes in a 3D representation and synthesizes new views through radiance fields. Since NeRF relies on the network weights to store the scene's 3-dimensional representation, BioNeRF implements a cognitive-inspired mechanism that fuses inputs from multiple sources into a memory-like structure, improving the storing capacity and extracting more intrinsic and correlated information. BioNeRF also mimics a behavior observed in pyramidal cells concerning contextual information, in which the memory is provided as the context and combined with the inputs of two subsequent neural models, one responsible for producing the volumetric densities and the other the colors used to render the scene. Experimental results show that BioNeRF outperforms state-of-the-art results concerning a quality measure that encodes human perception in two datasets: real-world images and synthetic data.

cs.CV↗

arXiv2024

Adapting LLMs for the Medical Domain in Portuguese: A Study on Fine-Tuning and Model Evaluation

This study evaluates the performance of large language models (LLMs) as medical agents in Portuguese, aiming to develop a reliable and relevant virtual assistant for healthcare professionals. The HealthCareMagic-100k-en and MedQuAD datasets, translated from English using GPT-3.5, were used to fine-tune the ChatBode-7B model using the PEFT-QLoRA method. The InternLM2 model, with initial training on medical data, presented the best overall performance, with high precision and adequacy in metrics such as accuracy, completeness and safety. However, DrBode models, derived from ChatBode, exhibited a phenomenon of catastrophic forgetting of acquired medical knowledge. Despite this, these models performed frequently or even better in aspects such as grammaticality and coherence. A significant challenge was low inter-rater agreement, highlighting the need for more robust assessment protocols. This work paves the way for future research, such as evaluating multilingual models specific to the medical field, improving the quality of training data, and developing more consistent evaluation methodologies for the medical field.

cs.CL↗

arXiv2024

A Review of Deep Learning-based Approaches for Deepfake Content Detection

Recent advancements in deep learning generative models have raised concerns as they can create highly convincing counterfeit images and videos. This poses a threat to people's integrity and can lead to social instability. To address this issue, there is a pressing need to develop new computational models that can efficiently detect forged content and alert users to potential image and video manipulations. This paper presents a comprehensive review of recent studies for deepfake content detection using deep learning-based approaches. We aim to broaden the state-of-the-art research by systematically reviewing the different categories of fake content detection. Furthermore, we report the advantages and drawbacks of the examined works, and prescribe several future directions towards the issues and shortcomings still unsolved on deepfake detection.

cs.CV↗

arXiv2024

Introducing Bode: A Fine-Tuned Large Language Model for Portuguese Prompt-Based Task

Large Language Models (LLMs) are increasingly bringing advances to Natural Language Processing. However, low-resource languages, those lacking extensive prominence in datasets for various NLP tasks, or where existing datasets are not as substantial, such as Portuguese, already obtain several benefits from LLMs, but not to the same extent. LLMs trained on multilingual datasets normally struggle to respond to prompts in Portuguese satisfactorily, presenting, for example, code switching in their responses. This work proposes a fine-tuned LLaMA 2-based model for Portuguese prompts named Bode in two versions: 7B and 13B. We evaluate the performance of this model in classification tasks using the zero-shot approach with in-context learning, and compare it with other LLMs. Our main contribution is to bring an LLM with satisfactory results in the Portuguese language, as well as to provide a model that is free for research or commercial purposes.

cs.CL↗

arXiv2023

Feature Selection and Hyperparameter Fine-tuning in Artificial Neural Networks for Wood Quality Classification

Quality classification of wood boards is an essential task in the sawmill industry, which is still usually performed by human operators in small to median companies in developing countries. Machine learning algorithms have been successfully employed to investigate the problem, offering a more affordable alternative compared to other solutions. However, such approaches usually present some drawbacks regarding the proper selection of their hyperparameters. Moreover, the models are susceptible to the features extracted from wood board images, which influence the induction of the model and, consequently, its generalization power. Therefore, in this paper, we investigate the problem of simultaneously tuning the hyperparameters of an artificial neural network (ANN) as well as selecting a subset of characteristics that better describes the wood board quality. Experiments were conducted over a private dataset composed of images obtained from a sawmill industry and described using different feature descriptors. The predictive performance of the model was compared against five baseline methods as well as a random search, performing either ANN hyperparameter tuning and feature selection. Experimental results suggest that hyperparameters should be adjusted according to the feature set, or the features should be selected considering the hyperparameter values. In summary, the best predictive performance, i.e., a balanced accuracy of $0.80$, was achieved in two distinct scenarios: (i) performing only feature selection, and (ii) performing both tasks concomitantly. Thus, we suggest that at least one of the two approaches should be considered in the context of industrial applications.

cs.LG↗

arXiv2023

Facial Point Graphs for Amyotrophic Lateral Sclerosis Identification

Identifying Amyotrophic Lateral Sclerosis (ALS) in its early stages is essential for establishing the beginning of treatment, enriching the outlook, and enhancing the overall well-being of those affected individuals. However, early diagnosis and detecting the disease's signs is not straightforward. A simpler and cheaper way arises by analyzing the patient's facial expressions through computational methods. When a patient with ALS engages in specific actions, e.g., opening their mouth, the movement of specific facial muscles differs from that observed in a healthy individual. This paper proposes Facial Point Graphs to learn information from the geometry of facial images to identify ALS automatically. The experimental outcomes in the Toronto Neuroface dataset show the proposed approach outperformed state-of-the-art results, fostering promising developments in the area.

cs.CV↗

arXiv2023

Canonical Cortical Graph Neural Networks and its Application for Speech Enhancement in Audio-Visual Hearing Aids

Despite the recent success of machine learning algorithms, most models face drawbacks when considering more complex tasks requiring interaction between different sources, such as multimodal input data and logical time sequences. On the other hand, the biological brain is highly sharpened in this sense, empowered to automatically manage and integrate such streams of information. In this context, this work draws inspiration from recent discoveries in brain cortical circuits to propose a more biologically plausible self-supervised machine learning approach. This combines multimodal information using intra-layer modulations together with Canonical Correlation Analysis, and a memory mechanism to keep track of temporal data, the overall approach termed Canonical Cortical Graph Neural networks. This is shown to outperform recent state-of-the-art models in terms of clean audio reconstruction and energy efficiency for a benchmark audio-visual speech dataset. The enhanced performance is demonstrated through a reduced and smother neuron firing rate distribution. suggesting that the proposed model is amenable for speech enhancement in future audio-visual hearing aid devices.

cs.SD↗

arXiv2022

Extractive Text Summarization Using Generalized Additive Models with Interactions for Sentence Selection

Automatic Text Summarization (ATS) is becoming relevant with the growth of textual data; however, with the popularization of public large-scale datasets, some recent machine learning approaches have focused on dense models and architectures that, despite producing notable results, usually turn out in models difficult to interpret. Given the challenge behind interpretable learning-based text summarization and the importance it may have for evolving the current state of the ATS field, this work studies the application of two modern Generalized Additive Models with interactions, namely Explainable Boosting Machine and GAMI-Net, to the extractive summarization problem based on linguistic features and binary classification.

cs.CL↗

arXiv2022

A survey on text generation using generative adversarial networks

This work presents a thorough review concerning recent studies and text generation advancements using Generative Adversarial Networks. The usage of adversarial learning for text generation is promising as it provides alternatives to generate the so-called "natural" language. Nevertheless, adversarial text generation is not a simple task as its foremost architecture, the Generative Adversarial Networks, were designed to cope with continuous information (image) instead of discrete data (text). Thus, most works are based on three possible options, i.e., Gumbel-Softmax differentiation, Reinforcement Learning, and modified training objectives. All alternatives are reviewed in this survey as they present the most recent approaches for generating text using adversarial-based techniques. The selected works were taken from renowned databases, such as Science Direct, IEEEXplore, Springer, Association for Computing Machinery, and arXiv, whereas each selected work has been critically analyzed and assessed to present its objective, methodology, and experimental results.

cs.CL↗

arXiv2022

Improving Pre-Trained Weights Through Meta-Heuristics Fine-Tuning

Machine Learning algorithms have been extensively researched throughout the last decade, leading to unprecedented advances in a broad range of applications, such as image classification and reconstruction, object recognition, and text categorization. Nonetheless, most Machine Learning algorithms are trained via derivative-based optimizers, such as the Stochastic Gradient Descent, leading to possible local optimum entrapments and inhibiting them from achieving proper performances. A bio-inspired alternative to traditional optimization techniques, denoted as meta-heuristic, has received significant attention due to its simplicity and ability to avoid local optimums imprisonment. In this work, we propose to use meta-heuristic techniques to fine-tune pre-trained weights, exploring additional regions of the search space, and improving their effectiveness. The experimental evaluation comprises two classification tasks (image and text) and is assessed under four literature datasets. Experimental results show nature-inspired algorithms' capacity in exploring the neighborhood of pre-trained weights, achieving superior results than their counterpart pre-trained architectures. Additionally, a thorough analysis of distinct architectures, such as Multi-Layer Perceptron and Recurrent Neural Networks, attempts to visualize and provide more precise insights into the most critical weights to be fine-tuned in the learning process.

cs.AI↗

arXiv2022

PL-kNN: A Parameterless Nearest Neighbors Classifier

Demands for minimum parameter setup in machine learning models are desirable to avoid time-consuming optimization processes. The $k$-Nearest Neighbors is one of the most effective and straightforward models employed in numerous problems. Despite its well-known performance, it requires the value of $k$ for specific data distribution, thus demanding expensive computational efforts. This paper proposes a $k$-Nearest Neighbors classifier that bypasses the need to define the value of $k$. The model computes the $k$ value adaptively considering the data distribution of the training set. We compared the proposed model against the standard $k$-Nearest Neighbors classifier and two parameterless versions from the literature. Experiments over 11 public datasets confirm the robustness of the proposed approach, for the obtained results were similar or even better than its counterpart versions.

cs.LG↗

arXiv2022

Multimodal Audio-Visual Information Fusion using Canonical-Correlated Graph Neural Network for Energy-Efficient Speech Enhancement

This paper proposes a novel multimodal self-supervised architecture for energy-efficient audio-visual (AV) speech enhancement that integrates Graph Neural Networks with canonical correlation analysis (CCA-GNN). The proposed approach lays its foundations on a state-of-the-art CCA-GNN that learns representative embeddings by maximizing the correlation between pairs of augmented views of the same input while decorrelating disconnected features. The key idea of the conventional CCA-GNN involves discarding augmentation-variant information and preserving augmentation-invariant information while preventing capturing of redundant information. Our proposed AV CCA-GNN model deals with multimodal representation learning context. Specifically, our model improves contextual AV speech processing by maximizing canonical correlation from augmented views of the same channel and canonical correlation from audio and visual embeddings. In addition, it proposes a positional node encoding that considers a prior-frame sequence distance instead of a feature-space representation when computing the node's nearest neighbors, introducing temporal information in the embeddings through the neighborhood's connectivity. Experiments conducted on the benchmark ChiME3 dataset show that our proposed prior frame-based AV CCA-GNN ensures better feature learning in the temporal context, leading to more energy-efficient speech reconstruction than state-of-the-art CCA-GNN and multilayer perceptron.

cs.SD↗