Large Language Models achieve remarkable performance but incur substantial computational costs unsuitable for resource-constrained deployments. This paper presents the first comprehensive task-specific efficiency analysis comparing 16 language models across five diverse NLP tasks. We introduce the Performance-Efficiency Ratio (PER), a novel metric integrating accuracy, throughput, memory, and latency through geometric mean normalization. Our systematic evaluation reveals that small models (0.5--3B parameters) achieve superior PER scores across all given tasks. These findings establish quantitative foundations for deploying small models in production environments prioritizing inference efficiency over marginal accuracy gains.
Interactive communication in virtual reality can be used in experimental paradigms to increase the ecological validity of hearing device evaluations. This requires the virtual environment to elicit natural communication behaviour in listeners. This study evaluates the effect of virtual animated characters' head movements on participants' communication behaviour and experience. Triadic conversations were conducted between a test participant and two confederates. To facilitate the manipulation of head movements, the conversation was conducted in telepresence using a system that transmitted audio, head movement data and video with low delay. The confederates were represented by virtual animated characters (avatars) with different levels of animation: Static heads, automated head movement animations based on speech level onsets, and animated head movements based on the transmitted head movements of the interlocutors. A condition was also included in which the videos of the interlocutors' heads were embedded in the visual scene. The results show significant effects of animation level on the participants' speech and head movement behaviour as recorded by physical sensors, as well as on the subjective sense of presence and the success of the conversation. The largest effects were found for the range of head orientation during speech and the perceived realism of avatars. Participants reported that they were spoken to in a more helpful way when the avatars showed head movements transmitted from the interlocutors than when the avatars' heads were static. We therefore conclude that the representation of interlocutors must include sufficiently realistic head movements in order to elicit natural communication behaviour.
Andrej Lúčny, Matilde Antonj, Carlo Mazzola
et al.
Human--robot interaction requires robots whose actions are legible, allowing humans to interpret, predict, and feel safe around them. This study investigates the legibility of humanoid robot arm movements in a pointing task, aiming to understand how humans predict robot intentions from truncated movements and bodily cues. We designed an experiment using the NICO humanoid robot, where participants observed its arm movements towards targets on a touchscreen. Robot cues varied across conditions: gaze, pointing, and pointing with congruent or incongruent gaze. Arm trajectories were stopped at 60\% or 80\% of their full length, and participants predicted the final target. We tested the multimodal superiority and ocular primacy hypotheses, both of which were supported by the experiment.
We study the uniqueness and regularity of minimizing movements solutions of a droplet model in the case of piecewise monotone forcing. We show that such solutions evolve uniquely on each interval of monotonicity, but branching non-uniqueness may occur where jumps and monotonicity changes coincide. This classification of minimizing movements solutions allows us to reduce the quasi-static evolution to a finite sequence of elliptic problems and establish $L^\infty_tC^{1,1/2-}_x$-regularity of solutions.
We demonstrate the feasibility of the radar-based measurement of body movements in scenarios involving multiple students using a pair of 79-GHz millimeter-wave radar systems with array antennas. We quantify the body motion using the Doppler frequency calculated from radar echoes. The measurement accuracy is evaluated for two experimental scenarios, namely university students in an office and elementary school students in a classroom. The body movements measured using the two radar systems are compared to evaluate the repeatability and angle dependency of the measurement. Moreover, in the first scenario, we compare the radar-estimated body movement with subjective evaluation scores provided by two evaluators. In the first scenario, the coefficient of correlation between the radar-estimated body movement and the subjective evaluation score is 0.73 on average, with a maximum value of 0.97; in the second scenario, the average correlation coefficient of body movements measured using two radar systems is as high as 0.78. These results indicate that the proposed approach can be used to monitor the body movements of multiple students in realistic scenarios.
The task of music-driven dance generation involves creating coherent dance movements that correspond to the given music. While existing methods can produce physically plausible dances, they often struggle to generalize to out-of-set data. The challenge arises from three aspects: 1) the high diversity of dance movements and significant differences in the distribution of music modalities, which make it difficult to generate music-aligned dance movements. 2) the lack of a large-scale music-dance dataset, which hinders the generation of generalized dance movements from music. 3) The protracted nature of dance movements poses a challenge to the maintenance of a consistent dance style. In this work, we introduce the EnchantDance framework, a state-of-the-art method for dance generation. Due to the redundancy of the original dance sequence along the time axis, EnchantDance first constructs a strong dance latent space and then trains a dance diffusion model on the dance latent space. To address the data gap, we construct a large-scale music-dance dataset, ChoreoSpectrum3D Dataset, which includes four dance genres and has a total duration of 70.32 hours, making it the largest reported music-dance dataset to date. To enhance consistency between music genre and dance style, we pre-train a music genre prediction network using transfer learning and incorporate music genre as extra conditional information in the training of the dance diffusion model. Extensive experiments demonstrate that our proposed framework achieves state-of-the-art performance on dance quality, diversity, and consistency.
Andrew J Duncan, Aaron Reeves, George J Gunn
et al.
The Cattle Tracing System database is an online recording system for cattle births, deaths and between--herd movements in the United Kingdom. Although it has been thoroughly examined, the most recently reported movement analysis is from 2009. This article uses the database to construct weighted directed monthly movement networks for two distinct periods of time, 2004--2006 and 2015--2017, to quantify by how much the underlying structure of the network has changed. Substantial changes in network structure may influence policy--makers directly or may influence models built upon the network data, and these in turn could impact policy--makers and their assessment of risk. Four general network measures are used (total number of nodes with movements, movements, births and deaths), in conjunction with network metrics to describe each monthly network. Two updates of the database were examined to determine by how much the movement data stored for a particular time period had been cleansed between updates. Statistical models show that there is a statistically significant effect of the time period (2004--2006 vs 2015--2017) in the values of all network measures and six of nine network metrics. Changes in the sizes of both the Giant and Weakly Strongly Connected components predict reductions in the upper and lower bounds of the maximum epidemic size. Examination of the updates of the database show that there are differences in records between updates and therefore evidence of historical data changing between updates. Accurate modelling of disease spread through a network requires representative descriptions of the network. The authors recommend that where possible the most recent available data always be used for network modelling and that methods of network prediction be examined to mitigate for the time required for data to become available.
Este artículo permite conocer los documentos relacionados con la participación de países centroamericanos escritos en torno a las exposiciones universales del siglo XIX, con el ánimo de evidenciar su significancia como documentos reflejantes de la marcada intencionalidad por atraer la atención de potencias mundiales hacia Centroamérica, para lo cual, se recurrió, en la mayoría de las veces, a la reproducción discursiva y a la mimetización de la cultura occidental europea. Se analizan los casos de Costa Rica, Nicaragua, El Salvador y Guatemala, cada uno con las particularidades que los caracterizan.
This paper presents a neural network model to generate virtual violinist's 3-D skeleton movements from music audio. Improved from the conventional recurrent neural network models for generating 2-D skeleton data in previous works, the proposed model incorporates an encoder-decoder architecture, as well as the self-attention mechanism to model the complicated dynamics in body movement sequences. To facilitate the optimization of self-attention model, beat tracking is applied to determine effective sizes and boundaries of the training examples. The decoder is accompanied with a refining network and a bowing attack inference mechanism to emphasize the right-hand behavior and bowing attack timing. Both objective and subjective evaluations reveal that the proposed model outperforms the state-of-the-art methods. To the best of our knowledge, this work represents the first attempt to generate 3-D violinists' body movements considering key features in musical body movement.
El presente artículo hace referencia al aparato teórico y metodológico relativo al problema en la reproducción del discurso colonialista eurocéntrico de las élites centroamericanas en el contexto de las Grandes Exposiciones Universales del siglo XIX. Esta reproducción produjo una serie de tópicos, tensiones y tendencias discursivas que incidieron, y lo siguen haciendo en el presente, en las relaciones de desventaja económica, social y política del istmo centroamericano con respecto a las potencias occidentales noratlánticas.
El arribo de los españoles y la subsiguiente colonización del archipiélago de las Filipinas durante la segunda mitad del siglo XVI dan inicio a los viajes entre la ciudad de Manila y el puerto de Acapulco en México. La apertura de esta ruta marítima inaugura una época en la que, por vez primera, el comercio mundial se globaliza a través de estructuradas redes de intercambio. Previamente, los portugueses habían establecido una vía de comercio entre Europa, la India y China. Pero, con la ruta del galeón de Manila a través del Pacífico, Asia Oriental se vincula a América y se integra al flujo monetario generado por la producción de plata en Hispanoamérica. A cambio, gran cantidad de mercancías orientales inundan los puertos de las colonias españolas en América. A su vez, los comerciantes novohispanos adquieren un papel central en los intercambios entre México con Centroamérica y el Virreinato del Perú, al igual que con el Oriente por medio de la ruta del galeón de Manila. Este recorrido comercial prevalecería durante 250 años. También habrían de producirse numerosos flujos migratorios y la propagación de los cultivos de origen americano en el oriente asiático.
Proof assistants, such as Isabelle/HOL, offer tools to facilitate inductive theorem proving. Isabelle experts know how to use these tools effectively; however, they did not have a systematic way to encode their expertise. To address this problem, we present our domain-specific language, LiFtEr. LiFtEr allows experienced Isabelle users to encode their induction heuristics in a style independent of any problem domain. LiFtEr's interpreter mechanically checks if a given application of induction tool matches the heuristics specified by experienced users, thus systematically transferring experienced users' expertise to new Isabelle users.
We start by surveying the history of the idea of a fundamental conservation law and briefly examine the role conservation laws play in different classical contexts. In such contexts we find conservation laws to be useful, but often not essential. Next we consider the quantum setting, where the conceptual problems of the standard formalism obstruct a rigorous analysis of the issue. We then analyze the fate of energy conservation within the various viable paths to address such conceptual problems; in all cases we find no satisfactory way to define a (useful) notion of energy that is generically conserved. Finally, we focus on the implications of this for the semiclassical gravity program and conclude that Einstein's equations cannot be said to always hold.
Estamos ante una obra monumental de compilación y crítica de fuentes relativa a los tributarios de la Provincia de Chiapas a lo largo de todo el período colonial. La publicación comprende un tomo impreso muy voluminoso y una extensa base de datos disponible en la red. Sigue un resumen de los principales contenidos del tomo impreso.
Maarten Bieshaar, Malte Depping, Jan Schneegans
et al.
In near future, vulnerable road users (VRUs) such as cyclists and pedestrians will be equipped with smart devices and wearables which are capable to communicate with intelligent vehicles and other traffic participants. Road users are then able to cooperate on different levels, such as in cooperative intention detection for advanced VRU protection. Smart devices can be used to detect intentions, e.g., an occluded cyclist intending to cross the road, to warn vehicles of VRUs, and prevent potential collisions. This article presents a human activity recognition approach to detect the starting movement of cyclists wearing smart devices. We propose a novel two-stage feature selection procedure using a score specialized for robust starting detection reducing the false positive detections and leading to understandable and interpretable features. The detection is modelled as a classification problem and realized by means of a machine learning classifier. We introduce an auxiliary class, that models starting movements and allows to integrate early movement indicators, i.e., body part movements indicating future behaviour. In this way we improve the robustness and reduce the detection time of the classifier. Our empirical studies with real-world data originating from experiments which involve 49 test subjects and consists of 84 starting motions show that we are able to detect the starting movements early. Our approach reaches an F1-score of 67 % within 0.33 s after the first movement of the bicycle wheel. Investigations concerning the device wearing location show that for devices worn in the trouser pocket the detector has less false detections and detects starting movements faster on average. We found that we can further improve the results when we train distinct classifiers for different wearing locations.
In this paper, we present a novel method to recognize the types of crowd movement from crowd trajectories using agent-based motion models (AMMs). Our idea is to apply a number of AMMs, referred to as exemplar-AMMs, to describe the crowd movement. Specifically, we propose an optimization framework that filters out the unknown noise in the crowd trajectories and measures their similarity to the exemplar-AMMs to produce a crowd motion feature. We then address our real-world crowd movement recognition problem as a multi-label classification problem. Our experiments show that the proposed feature outperforms the state-of-the-art methods in recognizing both simulated and real-world crowd movements from their trajectories. Finally, we have created a synthetic dataset, SynCrowd, which contains 2D crowd trajectories in various scenarios, generated by various crowd simulators. This dataset can serve as a training set or benchmark for crowd analysis work.
Human movements are physical processes combining the classical mechanics of the human body moving in space and the biomechanics of the muscles generating the forces acting on the body under sophisticated sensory-motor control. One way to characterize movement performance is through measures of energy efficiency that relate the mechanical energy of the body and metabolic energy expended by the muscles. We expect the practical utility of such measures to be greater when human subjects execute movements that maximize energy efficiency. We therefore seek to understand if and when subjects select movements with that maximizing energy efficiency. We proceed using a model-based approach to describe movements which perform a task requiring the body to add or remove external mechanical work to or from an object. We use the specific example of walking gaits doing external mechanical work by pulling a cart, and estimate the relationship between the avg. walking speed and avg. step length. In the limit where no external work is done, we find that the estimated maximum energy efficiency walking gait is much slower than the walking gaits healthy adults typically select. We then modify the situation of the walking gait by introducing an idealized mechanical device that creates an adjustable mechanical advantage. The walking gaits that maximize the energy efficiency using the optimal mechanical advantage are again much slower than the walking gaits healthy adults typically select. We finally modify the situation so that the avg. walking speed is fixed and derive the pattern of the avg. step length and mechanical advantage that maximize energy efficiency.
Despite the importance of the variational principles of physics, there have been relatively few attempts to consider them for a realistic framework. In addition to the old teleological question, this paper continues the recent discussion regarding the modal involvement of the principle of least action and its relations with the Humean view of the laws of nature. The reality of possible paths in the principle of least action is examined from the perspectives of the contemporary metaphysics of modality and Leibniz's concept of essences or possibles striving for existence. I elaborate a modal interpretation of the principle of least action that replaces a classical representation of a system's motion along a single history in the actual modality by simultaneous motions along an infinite set of all possible histories in the possible modality. This model is based on an intuition that deep ontological connections exist between the possible paths in the principle of least action and possible quantum histories in the Feynman path integral. I interpret the action as a physical measure of the essence of every possible history. Therefore only one actual history has the highest degree of the essence and minimal action. To address the issue of necessity, I assume that the principle of least action has a general physical necessity and lies between the laws of motion with a limited physical necessity and certain laws with a metaphysical necessity.
The way data structures organize data is often a function of the sequence of past operations. The organization of data is referred to as the data structure's state, and the sequence of past operations constitutes the data structure's history. A data structure state can therefore be used as an oracle to derive information about its history. As a result, for history-sensitive applications, such as privacy in e-voting, incremental signature schemes, and regulatory compliant data retention; it is imperative to conceal historical information contained within data structure states. Data structure history can be hidden by making data structures history independent. In this paper, we explore how to achieve history independence. We observe that current history independence notions are significantly limited in number and scope. There are two existing notions of history independence -- weak history independence (WHI) and strong history independence (SHI). WHI does not protect against insider adversaries and SHI mandates canonical representations, resulting in inefficiency. We postulate the need for a broad, encompassing notion of history independence, which can capture WHI, SHI, and a broad spectrum of new history independence notions. To this end, we introduce $Δ$history independence ($Δ$HI), a generic game-based framework that is malleable enough to accommodate existing and new history independence notions. As an essential step towards formalizing $Δ$HI, we explore the concepts of abstract data types, data structures, machine models, memory representations and history independence. Finally, to bridge the gap between theory and practice, we outline a general recipe for building end-to-end, history independent systems and demonstrate the use of the recipe in designing two history independent file systems.