Nat Wannawas

Dr. Nat Wannawas


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

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

Representative Publications

View All Publications
10 September 2025 in IEEE Access
Verified
Journal Article
Open access
NWNat Wannawas
Nat Wannawas
JHJinpei Han
Jinpei Han
Jyotindra NarayanJyotindra Narayan
Jyotindra Narayan
BSBuntoeng Srikarun
Buntoeng Srikarun
Reinforcement learning (RL) with physics-informed dynamics models leverages known environments’ physics to improve data efficiency. Recently, we have introduced Gaussian Process Co-Adjustment (GPCA), a data-efficient model that conservatively refines simulator predictions, leading to the success of RL-controlled functional electrical stimulation (FES). However, GPCA’s behaviors and capabilities in broad settings remain unexplored. This study introduces context-aware GPCA (Ca-GPCA)—a novel extension of GPCA that incorporates a context encoder that learns latent representations (contexts) of transition histories. With context-augmented input, Ca-GPCA can generalize over heterogeneous datasets. Ca-GPCA consistently achieves 30% and 19% improvements in performance over GPCA in benchmark and FES control tasks, respectively, where the training data are collected from environments with different dynamics properties. These results demonstrate Ca-GPCA’s potential as a backbone for control systems that can operate instantly across diverse users and settings. In FES applications, for example, Ca-GPCA can utilize prior data from multiple individuals to perform in new users without requiring personalized calibration, paving the way for scalable, user-independent assistive technologies.

8 February 2026 in Robotics
Verified
Journal Article
Open access
PDF file
NWNat Wannawas
Nat Wannawas
Jyotindra NarayanJyotindra Narayan
Jyotindra Narayan
WNWarakom Nerdnoi
Warakom Nerdnoi
Arsanchai SukkueaArsanchai Sukkuea
Arsanchai Sukkuea
Functional Electrical Stimulation (FES) can restore motor function; however, achieving precise multi-joint control remains challenging due to nonlinear muscle dynamics and fatigue. Reinforcement Learning (RL) offers a promising solution, but practical deployment is hindered by the need for patient-specific calibration. This study investigates offline RL approaches for controlling planar arm movements using heterogeneous datasets, aiming to enable zero-shot transfer to new users. We develop a biomechanical arm model in MuJoCo and evaluate four RL algorithms coupled with three offline techniques: conservative Q learning (SAC-CQL and QBR-CQL), Randomized Ensemble (QBR-REM), and distributional RL (IQNBR). Across all conditions, IQNBR demonstrates robust learning and superior control performance, achieving an average RMSE of 3.8±0.6 cm, even when trained on mixed-quality data. These results highlight the potential of distributional RL as a base learning method to build generic FES controllers that can operate without exhaustive calibration, with broader implications for controlling robots with human-like actuation systems.

Verified
Journal Article
Open access
PDF file
Arsanchai SukkueaArsanchai Sukkuea
Arsanchai Sukkuea
PAPensiri Akkajit
Pensiri Akkajit
NTNanda Kumar Tharu
Nanda Kumar Tharu
NWNat Wannawas
Nat Wannawas
Visual malware recognition classifies byte-plot images of executables and reports near-perfect accuracy, yet it is rarely evaluated the way a deployed detector would be. We present ByteScope, a framework that adds a cosine-margin family classifier, a supervised contrastive embedding for rejecting unknown families, and a class-balanced malware detection head calibrated at a fixed false-positive rate to an image backbone, together with a leakage-audited, open-world evaluation protocol, and we evaluate it on MaleVis against nine convolutional neural network (CNN), Vision Transformer and hybrid backbones with three seeds. A perceptual-hash audit, confirmed by pixel-level, embedding-based and visual inspection, shows that 50.6% of the test images and 70.0% of the malware have a same-class near-duplicate in the training split. On Swin-T, ByteScope reaches the highest mean 26-class accuracy (89.3%) and de-duplicated accuracy (78.4%), matches the best baseline in unknown-family detection (area under the receiver operating characteristic curve, AUROC, 0.925 against 0.922), and halves the evasion rate of its backbone under image-domain benign padding at a validation-calibrated threshold (15.2% against 36.0%) by making the operating point more stable, an effect that we trace to the normalized cosine classifier; at matched test false-positive rates, robustness is similar. At a 1% false-positive rate, neither ByteScope nor any of the nine baselines detects more than 3% of malware on average, whereas several variants of our hybrid CNN–Transformer architecture exceed this level (6.8 to 48.1%, with a large variance across seeds), and roughly 52 benign validation images cannot certify any rate below 5.6%. We release the code, the near-duplicate list and the splits.