Esther Luna Colombini

Rafael Figueiredo Prudencio, Marcos R. O. A. Maximo, Esther Luna Colombini

With the widespread adoption of deep learning, reinforcement learning (RL) has experienced a dramatic increase in popularity, scaling to previously intractable problems, such as playing complex games from pixel observations, sustaining conversations with humans, and controlling robotic agents. However, there is still a wide range of domains inaccessible to RL due to the high cost and danger of interacting with the environment. Offline RL is a paradigm that learns exclusively from static datasets of previously collected interactions, making it feasible to extract policies from large and diverse training datasets. Effective offline RL algorithms have a much wider range of applications than online RL, being particularly appealing for real-world applications, such as education, healthcare, and robotics. In this work, we contribute with a unifying taxonomy to classify offline RL methods. Furthermore, we provide a comprehensive review of the latest algorithmic breakthroughs in the field using a unified notation as well as a review of existing benchmarks' properties and shortcomings. Additionally, we provide a figure that summarizes the performance of each method and class of methods on different dataset properties, equipping researchers with the tools to decide which type of algorithm is best suited for the problem at hand and identify which classes of algorithms look the most promising. Finally, we provide our perspective on open problems and propose future research directions for this rapidly growing field.

Gabriel Oliveira dos Santos, Diego A. B. Moreira, Alef Iury Ferreira et al.

This work introduces CAPIVARA, a cost-efficient framework designed to enhance the performance of multilingual CLIP models in low-resource languages. While CLIP has excelled in zero-shot vision-language tasks, the resource-intensive nature of model training remains challenging. Many datasets lack linguistic diversity, featuring solely English descriptions for images. CAPIVARA addresses this by augmenting text data using image captioning and machine translation to generate multiple synthetic captions in low-resource languages. We optimize the training pipeline with LiT, LoRA, and gradient checkpointing to alleviate the computational cost. Through extensive experiments, CAPIVARA emerges as state of the art in zero-shot tasks involving images and Portuguese texts. We show the potential for significant improvements in other low-resource languages, achieved by fine-tuning the pre-trained multilingual CLIP using CAPIVARA on a single GPU for 2 hours. Our model and code is available at https://github.com/hiaac-nlp/CAPIVARA.

Letícia Berto, Paula Costa, Alexandre Simões et al.

Humans have needs motivating their behavior according to intensity and context. However, we also create preferences associated with each action's perceived pleasure, which is susceptible to changes over time. This makes decision-making more complex, requiring learning to balance needs and preferences according to the context. To understand how this process works and enable the development of robots with a motivational-based learning model, we computationally model a motivation theory proposed by Hull. In this model, the agent (an abstraction of a mobile robot) is motivated to keep itself in a state of homeostasis. We added hedonic dimensions to see how preferences affect decision-making, and we employed reinforcement learning to train our motivated-based agents. We run three agents with energy decay rates representing different metabolisms in two different environments to see the impact on their strategy, movement, and behavior. The results show that the agent learned better strategies in the environment that enables choices more adequate according to its metabolism. The use of pleasure in the motivational mechanism significantly impacted behavior learning, mainly for slow metabolism agents. When survival is at risk, the agent ignores pleasure and equilibrium, hinting at how to behave in harsh scenarios.

Leonardo de Lellis Rossi, Leticia Mara Berto, Eric Rohmer et al.

The ability to automatically learn movements and behaviors of increasing complexity is a long-term goal in autonomous systems. Indeed, this is a very complex problem that involves understanding how knowledge is acquired and reused by humans as well as proposing mechanisms that allow artificial agents to reuse previous knowledge. Inspired by Jean Piaget's theory's first three sensorimotor substages, this work presents a cognitive agent based on CONAIM (Conscious Attention-Based Integrated Model) that can learn procedures incrementally. Throughout the paper, we show the cognitive functions required in each substage and how adding new functions helps address tasks previously unsolved by the agent. Experiments were conducted with a humanoid robot in a simulated environment modeled with the Cognitive Systems Toolkit (CST) performing an object tracking task. The system is modeled using a single procedural learning mechanism based on Reinforcement Learning. The increasing agent's cognitive complexity is managed by adding new terms to the reward function for each learning phase. Results show that this approach is capable of solving complex tasks incrementally.

Elton Cardoso do Nascimento, Alexandre da Silva Simões, Esther Luna Colombini et al.

World models represent a paradigm shift in generative AI, pursuing predictive understanding and controllable simulation of environments in a structured and generalizable way. We present World Machine, a generative world-modeling architecture for time series. It is a transformer-based architecture with latent states that enables adaptation to different amounts of observed data and contexts. This shows an improvement over traditional transformers, which have a computational and memory cost that scales quadratically with the context. Experiments on a proposed synthetic dataset, Toy1D, validate the approach's feasibility, demonstrate capabilities not found in conventional transformers, and highlight the contributions of each component of the training protocol.

Hudson M. S. Bruno, Esther L. Colombini, Sidney N. Givigi

Visual Simultaneous Localization and Mapping (VSLAM) is a fundamental technology for robotics applications. While VSLAM research has achieved significant advancements, its robustness under challenging situations, such as poor lighting, dynamic environments, motion blur, and sensor failures, remains a challenging issue. To address these challenges, we introduce a novel RGB-D dataset designed for evaluating the robustness of VSLAM systems. The dataset comprises real-world indoor scenes with dynamic objects, motion blur, and varying illumination, as well as emulated camera failures, including lens dirt, condensation, underexposure, and overexposure. Additionally, we offer open-source scripts for injecting camera failures into any images, enabling further customization by the research community. Our experiments demonstrate that ORB-SLAM2, a traditional VSLAM algorithm, and TartanVO, a Deep Learning-based VO algorithm, can experience performance degradation under these challenging conditions. Therefore, this dataset and the camera failure open-source tools provide a valuable resource for developing more robust VSLAM systems capable of handling real-world challenges.

Gabriel Oliveira dos Santos, Esther Luna Colombini, Sandra Avila

Automatically describing images using natural sentences is an important task to support visually impaired people's inclusion onto the Internet. It is still a big challenge that requires understanding the relation of the objects present in the image and their attributes and actions they are involved in. Then, visual interpretation methods are needed, but linguistic models are also necessary to verbally describe the semantic relations. This problem is known as Image Captioning. Although many datasets were proposed in the literature, the majority contains only English captions, whereas datasets with captions described in other languages are scarce. Recently, a movement called PraCegoVer arose on the Internet, stimulating users from social media to publish images, tag #PraCegoVer and add a short description of their content. Thus, inspired by this movement, we have proposed the #PraCegoVer, a multi-modal dataset with Portuguese captions based on posts from Instagram. It is the first large dataset for image captioning in Portuguese with freely annotated images. Further, the captions in our dataset bring additional challenges to the problem: first, in contrast to popular datasets such as MS COCO Captions, #PraCegoVer has only one reference to each image; also, both mean and variance of our reference sentence length are significantly greater than those in the MS COCO Captions. These two characteristics contribute to making our dataset interesting due to the linguistic aspect and the challenges that it introduces to the image captioning problem. We publicly-share the dataset at https://github.com/gabrielsantosrv/PraCegoVer.

Gabriel Moraes Barros, Esther Luna Colombini

Unmanned Aerial Vehicles (UAVs), or drones, have recently been used in several civil application domains from organ delivery to remote locations to wireless network coverage. These platforms, however, are naturally unstable systems for which many different control approaches have been proposed. Generally based on classic and modern control, these algorithms require knowledge of the robot's dynamics. However, recently, model-free reinforcement learning has been successfully used for controlling drones without any prior knowledge of the robot model. In this work, we present a framework to train the Soft Actor-Critic (SAC) algorithm to low-level control of a quadrotor in a go-to-target task. All experiments were conducted under simulation. With the experiments, we show that SAC can not only learn a robust policy, but it can also cope with unseen scenarios. Videos from the simulations are available in https://www.youtube.com/watch?v=9z8vGs0Ri5g and the code in https://github.com/larocs/SAC_uav.

Eduardo Y. Sakabe, Felipe S. Abrahão, Alexandre Simões et al.

Understanding and controlling the informational complexity of neural networks is a central challenge in machine learning, with implications for generalization, optimization, and model capacity. While most approaches rely on entropy-based loss functions and statistical metrics, these measures often fail to capture deeper, causally relevant algorithmic regularities embedded in network structure. We propose a shift toward algorithmic information theory, using Binarized Neural Networks (BNNs) as a first proxy. Grounded in algorithmic probability (AP) and the universal distribution it defines, our approach characterizes learning dynamics through a formal, causally grounded lens. We apply the Block Decomposition Method (BDM) -- a scalable approximation of algorithmic complexity based on AP -- and demonstrate that it more closely tracks structural changes during training than entropy, consistently exhibiting stronger correlations with training loss across varying model sizes and randomized training runs. These results support the view of training as a process of algorithmic compression, where learning corresponds to the progressive internalization of structured regularities. In doing so, our work offers a principled estimate of learning progression and suggests a framework for complexity-aware learning and regularization, grounded in first principles from information theory, complexity, and computability.

Alana Santana, Paula P. Costa, Esther L. Colombini

Autonomous artificial agents must be able to learn behaviors in complex environments without humans to design tasks and rewards. Designing these functions for each environment is not feasible, thus, motivating the development of intrinsic reward functions. In this paper, we propose using several cognitive elements that have been neglected for a long time to build an internal world model for an intrinsically motivated agent. Our agent performs satisfactory iterations with the environment, learning complex behaviors without needing previously designed reward functions. We used 18 Atari games to evaluate what cognitive skills emerge in games that require reactive and deliberative behaviors. Our results show superior performance compared to the state-of-the-art in many test cases with dense and sparse rewards.

Jesus Paucar-Escalante, Matheus Alves da Silva, Bruno De Lima Sanches et al.

Parkinson's disease (PD) is a neurological disorder requiring early and accurate diagnosis for effective management. Machine learning (ML) has emerged as a powerful tool to enhance PD classification and diagnostic accuracy, particularly by leveraging wearable sensor data. This survey comprehensively reviews current ML methodologies used in classifying Parkinsonian tremors, evaluating various tremor data acquisition methodologies, signal preprocessing techniques, and feature selection methods across time and frequency domains, highlighting practical approaches for tremor classification. The survey explores ML models utilized in existing studies, ranging from traditional methods such as Support Vector Machines (SVM) and Random Forests to advanced deep learning architectures like Convolutional Neural Networks (CNN) and Long Short-Term Memory networks (LSTM). We assess the efficacy of these models in classifying tremor patterns associated with PD, considering their strengths and limitations. Furthermore, we discuss challenges and discrepancies in current research and broader challenges in applying ML to PD diagnosis using wearable sensor data. We also outline future research directions to advance ML applications in PD diagnostics, providing insights for researchers and practitioners.

Raysa Benatti, Fabiana Severi, Sandra Avila et al.

Data derived from the realm of the social sciences is often produced in digital text form, which motivates its use as a source for natural language processing methods. Researchers and practitioners have developed and relied on artificial intelligence techniques to collect, process, and analyze documents in the legal field, especially for tasks such as text summarization and classification. While increasing procedural efficiency is often the primary motivation behind natural language processing in the field, several works have proposed solutions for human rights-related issues, such as assessment of public policy and institutional social settings. One such issue is the presence of gender biases in court decisions, which has been largely studied in social sciences fields; biased institutional responses to gender-based violence are a violation of international human rights dispositions since they prevent gender minorities from accessing rights and hamper their dignity. Natural language processing-based approaches can help detect these biases on a larger scale. Still, the development and use of such tools require researchers and practitioners to be mindful of legal and ethical aspects concerning data sharing and use, reproducibility, domain expertise, and value-charged choices. In this work, we (a) present an experimental framework developed to automatically detect gender biases in court decisions issued in Brazilian Portuguese and (b) describe and elaborate on features we identify to be critical in such a technology, given its proposed use as a support tool for research and assessment of court~activity.

Alana Santana, Esther Colombini

Attention is a state of arousal capable of dealing with limited processing bottlenecks in human beings by focusing selectively on one piece of information while ignoring other perceptible information. For decades, concepts and functions of attention have been studied in philosophy, psychology, neuroscience, and computing. Currently, this property has been widely explored in deep neural networks. Many different neural attention models are now available and have been a very active research area over the past six years. From the theoretical standpoint of attention, this survey provides a critical analysis of major neural attention models. Here we propose a taxonomy that corroborates with theoretical aspects that predate Deep Learning. Our taxonomy provides an organizational structure that asks new questions and structures the understanding of existing attentional mechanisms. In particular, 17 criteria derived from psychology and neuroscience classic studies are formulated for qualitative comparison and critical analysis on the 51 main models found on a set of more than 650 papers analyzed. Also, we highlight several theoretical issues that have not yet been explored, including discussions about biological plausibility, highlight current research trends, and provide insights for the future.

Gabriel Oliveira dos Santos, Esther Luna Colombini, Sandra Avila

This paper shows that CIDEr-D, a traditional evaluation metric for image description, does not work properly on datasets where the number of words in the sentence is significantly greater than those in the MS COCO Captions dataset. We also show that CIDEr-D has performance hampered by the lack of multiple reference sentences and high variance of sentence length. To bypass this problem, we introduce CIDEr-R, which improves CIDEr-D, making it more flexible in dealing with datasets with high sentence length variance. We demonstrate that CIDEr-R is more accurate and closer to human judgment than CIDEr-D; CIDEr-R is more robust regarding the number of available references. Our results reveal that using Self-Critical Sequence Training to optimize CIDEr-R generates descriptive captions. In contrast, when CIDEr-D is optimized, the generated captions' length tends to be similar to the reference length. However, the models also repeat several times the same word to increase the sentence length.

Iury Cleveston, Esther L. Colombini

Building vehicles capable of operating without human supervision requires the determination of the agent's pose. Visual Odometry (VO) algorithms estimate the egomotion using only visual changes from the input images. The most recent VO methods implement deep-learning techniques using convolutional neural networks (CNN) extensively, which add a substantial cost when dealing with high-resolution images. Furthermore, in VO tasks, more input data does not mean a better prediction; on the contrary, the architecture may filter out useless information. Therefore, the implementation of computationally efficient and lightweight architectures is essential. In this work, we propose the RAM-VO, an extension of the Recurrent Attention Model (RAM) for visual odometry tasks. RAM-VO improves the visual and temporal representation of information and implements the Proximal Policy Optimization (PPO) algorithm to learn robust policies. The results indicate that RAM-VO can perform regressions with six degrees of freedom from monocular input images using approximately 3 million parameters. In addition, experiments on the KITTI dataset demonstrate that RAM-VO achieves competitive results using only 5.7% of the available visual information.

Hudson M. S. Bruno, Esther L. Colombini

The Simultaneous Localization and Mapping (SLAM) problem addresses the possibility of a robot to localize itself in an unknown environment and simultaneously build a consistent map of this environment. Recently, cameras have been successfully used to get the environment's features to perform SLAM, which is referred to as visual SLAM (VSLAM). However, classical VSLAM algorithms can be easily induced to fail when either the motion of the robot or the environment is too challenging. Although new approaches based on Deep Neural Networks (DNNs) have achieved promising results in VSLAM, they still are unable to outperform traditional methods. To leverage the robustness of deep learning to enhance traditional VSLAM systems, we propose to combine the potential of deep learning-based feature descriptors with the traditional geometry-based VSLAM, building a new VSLAM system called LIFT-SLAM. Experiments conducted on KITTI and Euroc datasets show that deep learning can be used to improve the performance of traditional VSLAM systems, as the proposed approach was able to achieve results comparable to the state-of-the-art while being robust to sensorial noise. We enhance the proposed VSLAM pipeline by avoiding parameter tuning for specific datasets with an adaptive approach while evaluating how transfer learning can affect the quality of the features extracted.

Alana de Santana Correia, Esther Luna Colombini

In humans, Attention is a core property of all perceptual and cognitive operations. Given our limited ability to process competing sources, attention mechanisms select, modulate, and focus on the information most relevant to behavior. For decades, concepts and functions of attention have been studied in philosophy, psychology, neuroscience, and computing. For the last six years, this property has been widely explored in deep neural networks. Currently, the state-of-the-art in Deep Learning is represented by neural attention models in several application domains. This survey provides a comprehensive overview and analysis of developments in neural attention models. We systematically reviewed hundreds of architectures in the area, identifying and discussing those in which attention has shown a significant impact. We also developed and made public an automated methodology to facilitate the development of reviews in the area. By critically analyzing 650 works, we describe the primary uses of attention in convolutional, recurrent networks and generative models, identifying common subgroups of uses and applications. Furthermore, we describe the impact of attention in different application domains and their impact on neural networks' interpretability. Finally, we list possible trends and opportunities for further research, hoping that this review will provide a succinct overview of the main attentional models in the area and guide researchers in developing future approaches that will drive further improvements.

Esther L. Colombini, Alexandre S. Simões, Carlos H. C. Ribeiro

The information available to robots in real tasks is widely distributed both in time and space, requiring the agent to search for relevant data. In humans, that face the same problem when sounds, images and smells are presented to their sensors in a daily scene, a natural system is applied: Attention. As vision plays an important role in our routine, most research regarding attention has involved this sensorial system and the same has been replicated to the robotics field. However,most of the robotics tasks nowadays do not rely only in visual data, that are still costly. To allow the use of attentive concepts with other robotics sensors that are usually used in tasks such as navigation, self-localization, searching and mapping, a generic attentional model has been previously proposed. In this work, feature mapping functions were designed to build feature maps to this attentive model from data from range scanner and sonar sensors. Experiments were performed in a high fidelity simulated robotics environment and results have demonstrated the capability of the model on dealing with both salient stimuli and goal-driven attention over multiple features extracted from multiple sensors.