Path patterns

Wang, He; Ondřej, Jan; O’Sullivan, Carol

doi:10.1145/2856400.2856410

Cited by 23 publications

(5 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The sharp rise in interest of ML techniques and recent advancements in the field have sparked numerous endeavors in data visualization research. These initiatives aim to comprehend the inner workings of these models and ascertain the optimal approach for specific tasks through fair measures and effective data visualization techniques such as [13], [14]. Noteworthy contributions include [15]-a versatile solution facilitating comparison among diverse ML models by analyzing various subsets of a shared training and test dataset.…”

Section: Related Workmentioning

confidence: 99%

Enhancing Federated Learning Evaluation: Exploring Instance-Level Insights with SQUARES in Image Classification Models

Mayank Jindal

2024

jes

View full text Add to dashboard Cite

Federated Learning (FL) presents a novel approach within the domain of Machine Learning (ML)—enabling the training of ML models in a distributed manner. This paradigmatic shift ensures the preservation of user privacy by conducting the training process at the edge, where data remains localized on individual devices. This stands in contrast to the conventional centralized paradigm—where users transmit data to a central server for processing. The FL framework operates within a more heterogeneous environment characterized by diverse data distributions across clients, a departure from the engineer-analyzed datasets typical of centralized paradigms. Whereas, evaluation of FL models typically relies on statistical techniques such as accuracy, recall, precision, log loss, and the confusion matrix, alongside visualization methods. While these techniques provide an overview of the model's performance and data utilization—they may lack granularity when comparing models with disparate characteristics. However, the SQUARES technique, offers a more nuanced evaluation of model performance at the instance level. This approach facilitates the examination of data biases, outlier detection, and model behavior during training on individual samples. Therefore, this study presents the development and evaluation of an FL image classification model across various scenarios—utilizing the SQUARES prototype. In addressing these more intricate scenarios, we aim to augment traditional visualizations and metrics, thereby uncovering insights and nuances that may elude standard evaluation methods prevalent in ML benchmarks.

show abstract

Section: Related Workmentioning

confidence: 99%

Enhancing Federated Learning Evaluation: Exploring Instance-Level Insights with SQUARES in Image Classification Models

Mayank Jindal

2024

jes

View full text Add to dashboard Cite

show abstract

“…Existing research assumes certain principles on individual motions, for example minimum effort [GCC*10], power-law [KSG14], and so on, but they are based on simplified hypotheses. Real-world individual motions are almost always sub-optimal for the physics-based models, which significantly affects the visual realism of crowd animation [WOO16,HXZW20]. Therefore, we propose to advance the system in the human-solution subspace rather in the entire solution space.…”

Section: Human-solution Spacementioning

confidence: 99%

Model‐based Crowd Behaviours in Human‐solution Space

Wei

Wang

Zhang

et al. 2023

Computer Graphics Forum

Self Cite

View full text Add to dashboard Cite

Realistic crowd simulation has been pursued for decades, but it still necessitates tedious human labour and a lot of trial and error. The majority of currently used crowd modelling is either empirical (model‐based) or data‐driven (model‐free). Model‐based methods cannot fit observed data precisely, whereas model‐free methods are limited by the availability/quality of data and are uninterpretable. In this paper, we aim at taking advantage of both model‐based and data‐driven approaches. In order to accomplish this, we propose a new simulation framework built on a physics‐based model that is designed to be data‐friendly. Both the general prior knowledge about crowds encoded by the physics‐based model and the specific real‐world crowd data at hand jointly influence the system dynamics. With a multi‐granularity physics‐based model, the framework combines microscopic and macroscopic motion control. Each simulation step is formulated as an energy optimization problem, where the minimizer is the desired crowd behaviour. In contrast to traditional optimization‐based methods which seek the theoretical minimizer, we designed an acceleration‐aware data‐driven scheme to compute the minimizer from real‐world data in order to achieve higher realism by parameterizing both velocity and acceleration. Experiments demonstrate that our method can produce crowd animations that are more realistically behaved in a variety of scales and scenarios when compared to the earlier methods.

show abstract

“…Besides simulation, different metrics have been proposed to validate simulation fidelity, where comparing simulations with real data becomes popular [7,8,[30][31][32]. However, these methods are designed to compare two sets of 2D trajectories.…”

Section: Crowd Simulationmentioning

confidence: 99%

Crowd Simulation with Detailed Body Motion and Interaction

Yao¹,

Wang²,

Wang³

et al. 2022

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Crowd simulation methods generally focus on high fidelity 2D trajectories but ignore detailed 3D body animation which is normally added in a post-processing step. We argue that this is an intrinsic flaw as detailed body motions affect the 2D trajectories, especially when interactions are present between characters, and characters and the environment. In practice, this requires labor-intensive post-processing, fitting individual character animations onto simulated trajectories where anybody interactions need to be manually specified. In this paper, we propose a new framework to integrate the modeling of crowd motions with character motions, to enable their mutual influence, so that crowd simulation also incorporates agent-agent and agent-environment interactions. The whole framework is based on a three-level hierarchical control structure to effectively control the scene at different scales efficiently and consistently. To facilitate control, each character is modeled as an agent governed by four modules: visual system, blackboard system, decision system, and animation system. The animation system of the agent model consists of two modes: a traditional Finite State Machine (FSM) animation mode, and a motion matching mode. So an agent not only retains the flexibility of FSMs, but also has the advantage of motion matching which adapts detailed body movements for interactions with other agents and the environment. Our method is universal and applicable to most interaction scenarios in various environments in crowd animation, which cannot be achieved by prior work. We validate the fluency and realism of the proposed method by extensive experiments and user studies.

show abstract

Path patterns

Cited by 23 publications

References 47 publications

Enhancing Federated Learning Evaluation: Exploring Instance-Level Insights with SQUARES in Image Classification Models

Enhancing Federated Learning Evaluation: Exploring Instance-Level Insights with SQUARES in Image Classification Models

Model‐based Crowd Behaviours in Human‐solution Space

Crowd Simulation with Detailed Body Motion and Interaction

Contact Info

Product

Resources

About