Manuel Palermo

Manuel Palermo

Auto-generated profile, based on publicly available information

Manuel Palermo

Bosch Car Multimedia,Braga,Portugal

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 October 2022 in Scientific Data
Verified
Journal Article
Open access
PDF file
MPManuel Palermo
Manuel Palermo
JLJoão M. Lopes
João M. Lopes
JAJoão André
João André
AMAna C. Matias
Ana C. Matias
JCJoao Jose Cerqueira
Joao Jose Cerqueira
Cristina SantosCristina Santos
Cristina Santos
Monitoring gait and posture while using assisting robotic devices is relevant to attain effective assistance and assess the user’s progression throughout time. This work presents a multi-camera, multimodal, and detailed dataset involving 14 healthy participants walking with a wheeled robotic walker equipped with a pair of affordable cameras. Depth data were acquired at 30 fps and synchronized with inertial data from Xsens MTw Awinda sensors and kinematic data from the segments of the Xsens biomechanical model, acquired at 60 Hz. Participants walked with the robotic walker at 3 different gait speeds, across 3 different walking scenarios/paths at 3 different locations. In total, this dataset provides approximately 92 minutes of total recording time, which corresponds to nearly 166.000 samples of synchronized data. This dataset may contribute to the scientific research by allowing the development and evaluation of: (i) vision-based pose estimation algorithms, exploring classic or deep learning approaches; (ii) human detection and tracking algorithms; (iii) movement forecasting; and (iv) biomechanical analysis of gait/posture when using a rehabilitation device.

30 September 2022 in Scientific Data
Verified
Journal Article
Open access
PDF file
MPManuel Palermo
Manuel Palermo
SCSara M. Cerqueira
Sara M. Cerqueira
JAJoão André
João André
APAntónio Pereira
António Pereira
Cristina SantosCristina Santos
Cristina Santos
Wearable technology is expanding for motion monitoring. However, open challenges still limit its widespread use, especially in low-cost systems. Most solutions are either expensive commercial products or lower performance ad-hoc systems. Moreover, few datasets are available for the development of complete and general solutions. This work presents 2 datasets, with low-cost and high-end Magnetic, Angular Rate, and Gravity(MARG) sensor data. Provides data for the complete inertial pose pipeline analysis, starting from raw data, sensor-to-segment calibration, multi-sensor fusion, skeleton-kinematics, to complete Human pose. Contains data from 21 and 10 participants, respectively, performing 6 types of sequences, presenting high variability and complex dynamics with almost complete range-of-motion. Amounts to 3.5 M samples, synchronized with a ground-truth inertial motion capture system. Presents a method to evaluate data quality. This database may contribute to develop novel algorithms for each pipeline’s processing steps, with applications in inertial pose estimation algorithms, human movement forecasting, and motion assessment in industrial or rehabilitation settings. All data and code to process and analyze the complete pipeline is freely available.

7 December 2024
Verified
Conference Paper
Closed access
JDJoão Delgado
João Delgado
MPManuel Palermo
Manuel Palermo
DADomingos Araújo
Domingos Araújo
Dalila DurãesDalila Durães
Dalila Durães
Autonomous driving systems depend on various perception tasks to ensure a clear understanding of the road, requiring accuracy and speed. Traditionally, real-time performance was achieved by optimizing each task independently, resulting in inefficiencies when handling multiple tasks due to high memory and computation costs. Multi-Task Learning (MTL) has emerged as a solution to these issues by using generalized deep learning models that simultaneously infer multiple tasks, reducing Giga Floating Point Operations (GFLOPs), neural network size, and enhancing inference speeds. Additionally, when tasks share information, individual task performance can improve. This paper introduces a Multi-task Deep Learning model that performs both Semantic Segmentation and Lane Detection using LiDAR point clouds. By applying a Bird's Eye View transformation, the model predicts point-wise segmentation labels while detecting and clustering lane lines and curbsides into related instances. Experimental results demonstrate MTL's memory and time efficiency, halving resource usage and proving the feasibility of lane line detection solely from LiDAR data.