Search

Most of the prevalent approaches in speech prosody modeling rely on learning global style representations in a continuous latent space which encode and transfer the attributes of reference speech. However, recent work on neural codecs which are based on Residual Vector Quantization (RVQ) already shows great potential offering distinct advantages. We investigate the prosody modeling capabilities of the discrete space of such an RVQ-VAE model, modifying it to operate on the phoneme-level. We condition both the encoder and decoder of the model on linguistic representations and apply a global speaker embedding in order to factor out both phonetic and speaker information. We conduct an extensive set of investigations based on subjective experiments and objective measures to show that the phoneme-level discrete latent representations obtained this way achieves a high degree of disentanglement, capturing fine-grained prosodic information that is robust and transferable. The latent space turns out to have interpretable structure with its principal components corresponding to pitch and energy.

Super-Alfv\'enic jets, with kinetic energy densities significantly exceeding that of the solar wind, are commonly generated downstream of Earth's bow shock under both high and low beta plasma conditions. In this study, we present theoretical evidence that these enhanced kinetic energy flows are driven by firehose-unstable fluctuations and compressive heating within collisionless plasma environments. Using a fluid formalism that incorporates pressure anisotropy, we estimate that the downstream flow of a collisionless plasma shock can be accelerated by a factor of 2 to 4 following the compression and saturation of firehose instability. By analyzing quasi-parallel magnetosheath jets observed in situ by the Magnetospheric Multiscale (MMS) mission, we find that approximately 11\% of plasma measurements within these jets exhibit firehose-unstable fluctuations. Our findings offer an explanation for the distinctive generation of fast downstream flows in both low ($\beta<1$) and high ($\beta>1$) beta plasmas, and provide new evidence that kinetic processes are crucial for accurately describing the formation and evolution of magnetosheath jets., Comment: 12 pages, under review; comments welcome

In the present paper, we outline and expound the fundamental and novel qualitative-cum-philosophical premises, principles, ideas, concepts, constructions and results that originate from our ongoing research project of applying the conceptual panoply and the technical machinery of Abstract Differential Geometry (ADG) to various persistently outstanding issues in Quantum Gravity (QG) research. At the end of the paper, we discuss the potential philosophical repercussions of two possible future research routes that the main stream of our applications of ADG to QG may bifurcate towards in view of three independent, but overlapping, research papers that are currently under development., Comment: Corrected, updated and slightly expanded Version II of the paper, to replace Version I

The celebrated $3x+1$ problem is reformulated via the use of an analytic expression of the trailing zeros sequence resulting in a single branch formula $f(x)+1$ with a unique fixed point. The resultant formula $f(x)$ is also found to coincide with that of the discrete derivative of the sorted sequence of fixed points of the reflection operator on even binary palindromes of fixed even length \textit{2k} in any interval $[0\cdots2^{2k}-1]$. A set of equivalent reformulations of the problem are also presented., Comment: 20 pages, 5 figures

Digital twin networks (DTNs) serve as an emerging facilitator in the industrial networking sector, enabling the management of new classes of services, which require tailored support for improved resource utilization, low latencies and accurate data fidelity. In this paper, we explore the intersection between theoretical recommendations and practical implications of applying DTNs to industrial networked environments, sharing empirical findings and lessons learned from our ongoing work. To this end, we first provide experimental examples from selected aspects of data representations and fidelity, mixed-criticality workload support, and application-driven services. Then, we introduce an architectural framework for DTNs, exposing a more practical extension of existing standards; notably the ITU-T Y.3090 (2022) recommendation. Specifically, we explore and discuss the dual nature of DTNs, meant as a digital twin of the network and a network of digital twins, allowing the co-existence of both paradigms., Comment: Funding: (i) European Union under the Italian National Recovery and Resilience Plan (NRRP) of NextGenerationEU, partnership on "Telecommunications of the Future" PE00000001 - program "RESTART", (ii) RA Science Committee grant No. 22rl-052

A new algebra, hitherto not encountered in the usual Lie algebraic varieties or supervarieties, is introduced. The paper explores the rich and novel structure of the algebra, and it compares it on the one hand with the Jordan-Lie Superalgebras studied by Okubo and Kamiya, and on the other, with the four usual Euclidean division rings of the reals (R), the complexes (C), the quaternions (H) and the octonions (O), key algebraic properties of which the algebra is seen to combine, alter, extend and generalise. A potential physical application of the algebra is briefly alluded to at the end., Comment: arXiv admin note: substantial text overlap with arXiv:math-ph/0110030

Emotion detection in textual data has received growing interest in recent years, as it is pivotal for developing empathetic human-computer interaction systems. This paper introduces a method for categorizing emotions from text, which acknowledges and differentiates between the diversified similarities and distinctions of various emotions. Initially, we establish a baseline by training a transformer-based model for standard emotion classification, achieving state-of-the-art performance. We argue that not all misclassifications are of the same importance, as there are perceptual similarities among emotional classes. We thus redefine the emotion labeling problem by shifting it from a traditional classification model to an ordinal classification one, where discrete emotions are arranged in a sequential order according to their valence levels. Finally, we propose a method that performs ordinal classification in the two-dimensional emotion space, considering both valence and arousal scales. The results show that our approach not only preserves high accuracy in emotion prediction but also significantly reduces the magnitude of errors in cases of misclassification.

Conventional methods for outdoor environment reconstruction rely predominantly on vision-based techniques like photogrammetry and LiDAR, facing limitations such as constrained coverage, susceptibility to environmental conditions, and high computational and energy demands. These challenges are particularly pronounced in applications like augmented reality navigation, especially when integrated with wearable devices featuring constrained computational resources and energy budgets. In response, this paper proposes a novel approach harnessing ambient wireless signals for outdoor environment reconstruction. By analyzing radio frequency (RF) data, the paper aims to deduce the environmental characteristics and digitally reconstruct the outdoor surroundings. Investigating the efficacy of selected deep learning (DL) techniques on the synthetic RF dataset WAIR-D, the study endeavors to address the research gap in this domain. Two DL-driven approaches are evaluated (convolutional U-Net and CLIP+ based on vision transformers), with performance assessed using metrics like intersection-over-union (IoU), Hausdorff distance, and Chamfer distance. The results demonstrate promising performance of the RF-based reconstruction method, paving the way towards lightweight and scalable reconstruction solutions., Comment: This work has been submitted to the IEEE for possible publication. Work partly supported by the RA Science Committee grant No. 22rl-052 (DISTAL) and the EU under Italian National Recovery and Resilience Plan of NextGenerationEU on "Telecommunications of the Future" (PE00000001 - program "RESTART")

We continue ongoing research work on applying the homological algebraic conceptual and technical machinery of Abstract Differential Geometry towards formulating a finitary, causal and quantal version of vacuum Einstein Lorentzian gravity and free Yang-Mills theories., Comment: This is an invited paper contribution to a Special Volume titled `Physical Geometry: Unravelling the Weave of Quantum Geometry' in memory and honour of Professor Anastasios Mallios, edited by Dr Elias Zafiris. Extended latest version includes two Addenda and an Appendix at the back

This paper deals with the problem of informative path planning for a UAV deployed for precision agriculture applications. First, we observe that the ``fear of missing out'' data lead to uniform, conservative scanning policies over the whole agricultural field. Consequently, employing a non-uniform scanning approach can mitigate the expenditure of time in areas with minimal or negligible real value, while ensuring heightened precision in information-dense regions. Turning to the available informative path planning methodologies, we discern that certain methods entail intensive computational requirements, while others necessitate training on an ideal world simulator. To address the aforementioned issues, we propose an active sensing coverage path planning approach, named OverFOMO, that regulates the speed of the UAV in accordance with both the relative quantity of the identified classes, i.e. crops and weeds, and the confidence level of such detections. To identify these instances, a robust Deep Learning segmentation model is deployed. The computational needs of the proposed algorithm are independent of the size of the agricultural field, rendering its applicability on modern UAVs quite straightforward. The proposed algorithm was evaluated with a simu-realistic pipeline, combining data from real UAV missions and the high-fidelity dynamics of AirSim simulator, showcasing its performance improvements over the established state of affairs for this type of missions. An open-source implementation of the algorithm and the evaluation pipeline is also available: \url{https://github.com/emmarapt/OverFOMO}.

Measurements of chemical abundances in high-$z$ star-forming (SF) galaxies place important constraints on the enrichment histories of galaxies and the physical conditions in the early universe. JWST is beginning to enable direct chemical abundance measurements in galaxies at $z$$>$2 via the detection of the faint T$_e$-sensitive auroral line [O III]$\lambda$4364. However, abundances of other elements (e.g., S and Ar) in high-$z$ galaxies remain unconstrained due to a lack of T$_e$ data and wavelength coverage. Here, we present multiple direct abundances in Q2343-D40, a galaxy at $z=$2.9628$\pm$0.0001 observed with JWST/NIRSpec as part of the CECILIA program. We report the first simultaneous measurement of T$_e$[O III] and T$_e$[S III] in a high-$z$ galaxy, finding good agreement with the temperature trends in local SF systems. We measure a gas-phase metallicity of 12+log(O/H) $=8.07\pm0.06$, and the N/O abundance, log(N/O) $=-1.37\pm0.21$, is indicative of primary nucleosynthesis. The S/O and Ar/O relative abundances, log(S/O)$=-1.88\pm0.10$ and log(Ar/O)$=-2.80\pm0.12$, are both $>$0.3 dex lower than the solar ratios. However, the relative Ar$^{2+}$/S$^{2+}$ abundance is consistent with the solar ratio, suggesting that the relative S-to-Ar abundance does not evolve significantly with redshift. Recent nucleosynthesis models find that a significant amount of S and Ar are produced in Type Ia supernovae, such that the S/O and Ar/O abundances in Q2343-D40 could be the result of predominantly core-collapse supernovae enrichment. Future JWST observations of high-$z$ galaxies will uncover whether S/O and Ar/O are sensitive to the timescales of these different enrichment mechanisms., Comment: 14 pages, 3 figures, accepted for publication in The Astrophysical Journal Letters

We present a computational modelling approach which targets capturing the specifics on how to virtually augment a Metaverse user's available social time capacity via using an independent and autonomous version of her digital representation in the Metaverse. We motivate why this is a fundamental building block to model large-scale social networks in the Metaverse, and emerging properties herein. We envision a Metaverse-focused extension of the traditional avatar concept: An avatar can be as well programmed to operate independently when its user is not controlling it directly, thus turning it into an agent-based digital human representation. This way, we highlight how such an independent avatar could help its user to better navigate their social relationships and optimize their socializing time in the Metaverse by (partly) offloading some interactions to the avatar. We model the setting and identify the characteristic variables by using selected concepts from social sciences: ego networks, social presence, and social cues. Then, we formulate the problem of maximizing the user's non-avatar-mediated spare time as a linear optimization. Finally, we analyze the feasible region of the problem and we present some initial insights on the spare time that can be achieved for different parameter values of the avatar-mediated interactions., Comment: Supported by the projects: Piano Nazionale di Ripresa e Resilienza IR0000013 - "SoBigData.it", Partenariato Esteso PE00000013 - "FAIR", Centro Nazionale CN00000013 - "ICSC"

We propose Hierarchical Text Spotter (HTS), a novel method for the joint task of word-level text spotting and geometric layout analysis. HTS can recognize text in an image and identify its 4-level hierarchical structure: characters, words, lines, and paragraphs. The proposed HTS is characterized by two novel components: (1) a Unified-Detector-Polygon (UDP) that produces Bezier Curve polygons of text lines and an affinity matrix for paragraph grouping between detected lines; (2) a Line-to-Character-to-Word (L2C2W) recognizer that splits lines into characters and further merges them back into words. HTS achieves state-of-the-art results on multiple word-level text spotting benchmark datasets as well as geometric layout analysis tasks., Comment: Accepted to WACV 2024

We present the first results from CECILIA, a Cycle 1 JWST NIRSpec/MSA program that uses ultra-deep ~30 hour G235M/F170LP observations to target multiple electron temperature-sensitive auroral lines in the spectra of 33 galaxies at z~1-3. Using a subset of 23 galaxies, we construct two ~600 object-hour composite spectra, both with and without the stellar continuum, and use these to investigate the characteristic rest-optical (5700-8500 Angstrom) spectrum of star-forming galaxies at the peak epoch of cosmic star formation. Emission lines of eight different elements (H, He, N, O, Si, S, Ar, and Ni) are detected, with most of these features observed to be <3% the strength of H-alpha. We report the characteristic strength of three auroral lines ([NII]5756, [SIII]6313, and [OII]7322,7332), as well as other semi-strong and faint emission lines, including forbidden [NiII]7380,7414 and the OI 8449 recombination line, some of which have never before been observed outside of the local universe. Using these measurements, we find T_e[NII]=13630+/-2540 K, representing the first measurement of electron temperature using [NII] in the high-redshift universe. We also see evidence for broad line emission with a FWHM of ~536 km/s; the broad component of H-alpha is 6.01-28.31% the strength of the narrow component and likely arises from star-formation driven outflows. Finally, we briefly comment on the feasibility of obtaining large samples of faint emission lines using JWST in the future., Comment: Published in ApJL in November 2023

We give a diagrammatic summary of the connections between various theorems and conjectures about the vanishing of the Euler characteristic., Comment: v3: corrected metadata poster URL; v2: small additions; 1 diagram, 6 pages; poster version at https://go.ur.de/euler-roadmap

We organize a competition on hierarchical text detection and recognition. The competition is aimed to promote research into deep learning models and systems that can jointly perform text detection and recognition and geometric layout analysis. We present details of the proposed competition organization, including tasks, datasets, evaluations, and schedule. During the competition period (from January 2nd 2023 to April 1st 2023), at least 50 submissions from more than 20 teams were made in the 2 proposed tasks. Considering the number of teams and submissions, we conclude that the HierText competition has been successfully held. In this report, we will also present the competition results and insights from them., Comment: ICDAR 2023 competition report by organizers (accepted and to be published officially later)

The challenging problem of non-line-of-sight (NLOS) localization is critical for many wireless networking applications. The lack of available datasets has made NLOS localization difficult to tackle with ML-driven methods, but recent developments in synthetic dataset generation have provided new opportunities for research. This paper explores three different input representations: (i) single wireless radio path features, (ii) wireless radio link features (multi-path), and (iii) image-based representations. Inspired by the two latter new representations, we design two convolutional neural networks (CNNs) and we demonstrate that, although not significantly improving the NLOS localization performance, they are able to support richer prediction outputs, thus allowing deeper analysis of the predictions. In particular, the richer outputs enable reliable identification of non-trustworthy predictions and support the prediction of the top-K candidate locations for a given instance. We also measure how the availability of various features (such as angles of signal departure and arrival) affects the model's performance, providing insights about the types of data that should be collected for enhanced NLOS localization. Our insights motivate future work on building more efficient neural architectures and input representations for improved NLOS localization performance, along with additional useful application features., Comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible. Work partly supported by the RA Science Committee grant No. 22rl-052 (DISTAL) and the EU under Italian National Recovery and Resilience Plan of NextGenerationEU on "Telecommunications of the Future" (PE00000001 - program "RESTART")

The cyber-physical convergence is opening up new business opportunities for industrial operators. The need for deep integration of the cyber and the physical worlds establishes a rich business agenda towards consolidating new system and network engineering approaches. This revolution would not be possible without the rich and heterogeneous sources of data, as well as the ability of their intelligent exploitation, mainly due to the fact that data will serve as a fundamental resource to promote Industry 4.0. One of the most fruitful research and practice areas emerging from this data-rich, cyber-physical, smart factory environment is the data-driven process monitoring field, which applies machine learning methodologies to enable predictive maintenance applications. In this paper, we examine popular time series forecasting techniques as well as supervised machine learning algorithms in the applied context of Industry 4.0, by transforming and preprocessing the historical industrial dataset of a packing machine's operational state recordings (real data coming from the production line of a manufacturing plant from the food and beverage domain). In our methodology, we use only a single signal concerning the machine's operational status to make our predictions, without considering other operational variables or fault and warning signals, hence its characterization as ``agnostic''. In this respect, the results demonstrate that the adopted methods achieve a quite promising performance on three targeted use cases.

A large part of the expressive speech synthesis literature focuses on learning prosodic representations of the speech signal which are then modeled by a prior distribution during inference. In this paper, we compare different prior architectures at the task of predicting phoneme level prosodic representations extracted with an unsupervised FVAE model. We use both subjective and objective metrics to show that normalizing flow based prior networks can result in more expressive speech at the cost of a slight drop in quality. Furthermore, we show that the synthesized speech has higher variability, for a given text, due to the nature of normalizing flows. We also propose a Dynamical VAE model, that can generate higher quality speech although with decreased expressiveness and variability compared to the flow based models., Comment: Submitted to ICASSP 2023

This paper proposes an Expressive Speech Synthesis model that utilizes token-level latent prosodic variables in order to capture and control utterance-level attributes, such as character acting voice and speaking style. Current works aim to explicitly factorize such fine-grained and utterance-level speech attributes into different representations extracted by modules that operate in the corresponding level. We show that the fine-grained latent space also captures coarse-grained information, which is more evident as the dimension of latent space increases in order to capture diverse prosodic representations. Therefore, a trade-off arises between the diversity of the token-level and utterance-level representations and their disentanglement. We alleviate this issue by first capturing rich speech attributes into a token-level latent space and then, separately train a prior network that given the input text, learns utterance-level representations in order to predict the phoneme-level, posterior latents extracted during the previous step. Both qualitative and quantitative evaluations are used to demonstrate the effectiveness of the proposed approach. Audio samples are available in our demo page., Comment: Submitted to ICASSP 2023

The gender of any voice user interface is a key element of its perceived identity. Recently, there has been increasing interest in interfaces where the gender is ambiguous rather than clearly identifying as female or male. This work addresses the task of generating novel gender-ambiguous TTS voices in a multi-speaker, multilingual setting. This is accomplished by efficiently sampling from a latent speaker embedding space using a proposed gender-aware method. Extensive objective and subjective evaluations clearly indicate that this method is able to efficiently generate a range of novel, diverse voices that are consistent and perceived as more gender-ambiguous than a baseline voice across all the languages examined. Interestingly, the gender perception is found to be robust across two demographic factors of the listeners: native language and gender. To our knowledge, this is the first systematic and validated approach that can reliably generate a variety of gender-ambiguous voices., Comment: Accepted to INTERSPEECH 2023

This paper presents a method for end-to-end cross-lingual text-to-speech (TTS) which aims to preserve the target language's pronunciation regardless of the original speaker's language. The model used is based on a non-attentive Tacotron architecture, where the decoder has been replaced with a normalizing flow network conditioned on the speaker identity, allowing both TTS and voice conversion (VC) to be performed by the same model due to the inherent linguistic content and speaker identity disentanglement. When used in a cross-lingual setting, acoustic features are initially produced with a native speaker of the target language and then voice conversion is applied by the same model in order to convert these features to the target speaker's voice. We verify through objective and subjective evaluations that our method can have benefits compared to baseline cross-lingual synthesis. By including speakers averaging 7.5 minutes of speech, we also present positive results on low-resource scenarios., Comment: Fundamental changes to the model described and experimental procedure

For a small $n$-category $\mathscr{C}$ and an $n$-topos $\mathscr{X}$, we study necessary and sufficient conditions for a functor $f \colon \mathscr{C} \to \mathscr{X}$ to determine a geometric morphism from $\mathscr{X}$ to the $n$-topos $\mathcal{P}(\mathscr{C})_n$ of presheaves on $\mathscr{C}$ for any $n \geq 1$. These results generalize and unify results of Lurie for $n=\infty$ and classical characterizations of flat functors (Diaconescu's theorem) for $n=1$. Interestingly, for $n=\infty$, our analogue of Diaconescu's theorem requires hypercompleteness. As an application, we show that the $\infty$-topos associated to an $n$-site behaves as an $n$-localic $\infty$-topos with respect to hypercomplete $\infty$-topoi., Comment: 27 pages

Soil texture is key information in agriculture for improving soil knowledge and crop performance, so the accurate mapping of this crucial feature is imperative for rationally planning cultivations and for targeting interventions. We studied the relationship between radioelements and soil texture in the Mezzano Lowland (Italy), a 189 $km^2$ agricultural plain investigated through a ded-icated airborne gamma-ray spectroscopy survey. The K and Th abundances were used to retrieve the clay and sand content by means of a multi-approach method. Linear (simple and multiple) and non-linear (machine learning algorithms with deep neural networks) predictive models were trained and tested adopting a 1:50,000 scale soil texture map. The comparison of these approaches highlighted that the non-linear model introduces significant improvements in the prediction of soil texture fractions. The predicted maps of the clay and of the sand content were compared with the regional soil maps. Although the macro-structures were equally present, the airborne gam-ma-ray data permits us shedding light on finer features. Map areas with higher clay content were coincident with paleo-channels crossing the Mezzano Lowland in Etruscan and Roman periods, confirmed by the hydrographic setting of historical maps and by the geo-morphological features of the study area., Comment: 18 pages, 9 figures, 3 tables, correspondence to Andrea Maino via email at maino@fe.infn.it

Smart city management is going through a remarkable transition, in terms of quality and diversity of services provided to the end-users. The stakeholders that deliver pervasive applications are now able to address fundamental challenges in the big data value chain, from data acquisition, data analysis and processing, data storage and curation, and data visualisation in real scenarios. Industry 4.0 is pushing this trend forward, demanding for servitization of products and data, also for the smart cities sector where humans, sensors and devices are operating in strict collaboration. The data produced by the ubiquitous devices must be processed quickly to allow the implementation of reactive services such as situational awareness, video surveillance and geo-localization, while always ensuring the safety and privacy of involved citizens. This paper proposes a modular architecture to (i) leverage innovative technologies for data acquisition, management and distribution (such as Apache Kafka and Apache NiFi), (ii) develop a multi-layer engineering solution for revealing valuable and hidden societal knowledge in smart cities environment, and (iii) tackle the main issues in tasks involving complex data flows and provide general guidelines to solve them. We derived some guidelines from an experimental setting performed together with leading industrial technical departments to accomplish an efficient system for monitoring and servitization of smart city assets, with a scalable platform that confirms its usefulness in numerous smart city use cases with different needs., Comment: This work was funded by the European Union's Horizon 2020 research and innovation programme MARVEL under grant agreement No 957337. This publication reflects the authors views only. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

We study the simplicial coalgebra of chains on a simplicial set with respect to three notions of weak equivalence. To this end, we construct three model structures on the category of reduced simplicial sets for any commutative ring R. The weak equivalences are given by: (1) an R-linearized version of categorical equivalences, (2) maps inducing an isomorphism on fundamental groups and an R-homology equivalence between universal covers, and (3) R-homology equivalences. Analogously, for any field F, we construct three model structures on the category of connected simplicial cocommutative F-coalgebras. The weak equivalences in this context are (1') maps inducing a quasi-isomorphism of dg algebras after applying the cobar functor, (2') maps inducing a quasi-isomorphism of dg algebras after applying a localized version of the cobar functor, and (3') quasi-isomorphisms. Building on previous work of Goerss in the context of (3)-(3'), we prove that, when F is algebraically closed, the simplicial F-coalgebra of chains defines a homotopically full and faithful left Quillen functor for each pair of model categories. More generally, when F is a perfect field, we compare the three pairs of model categories in terms of suitable notions of homotopy fixed points with respect to the absolute Galois group of F., Comment: Revised version following referee reports. The definition of the localized cobar construction has been modified so that the multiplicative set of elements to be inverted is extracted after applying the cobar functor. This makes the resulting construction invariant under Cobar-quasi-isomorphisms. To appear in IMRN

The recent advances in wireless energy transfer (WET) provide an alternate and reliable option for replenishing the battery of pervasive and portable devices, such as smartphones. The peer-to-peer (P2P) mode of WET brings improved flexibility to the charging process among the devices as they can maintain their mobility while replenishing their battery. Few existing works in P2P-WET unrealistically assume the nodes to be exchanging energy at every opportunity with any other node. Also, energy exchange between the nodes is not bounded by the energy transfer limit in that inter-node meeting duration. In this regard, the parametric heterogeneity (in terms of device's battery capacity and WET hardware) among the nodes also affects the energy transfer bound in each P2P interaction, and thus, may lead to unbalanced network energy distributions. This inherent heterogeneity aspect has not been adequately covered in the P2P-WET literature so far, especially from the point of view of maintaining a balanced energy distribution in the networked population. In this work, we present a Heterogeneity-aware Wireless Energy Transfer (HetWET) method. In contrast to the existing literature, we devise a fine-grained model of wireless energy transfer while considering the parametric heterogeneity of the participating devices. Thereafter, we enable the nodes to explore and dynamically decide the peers for energy exchange. The performance of HetWET is evaluated using extensive simulations with varying heterogeneity settings. The evaluation results demonstrate that HetWET can maintain lower energy losses and achieve more balanced energy variation distance compared to three different state-of-the-art methods., Comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible. This work was carried out during the tenure of an ERCIM 'Alain Bensoussan' Fellowship Programme of the first author