Fast and Accurate CNN‐based Brushing in Scatterplots

Fan, Chaoran; Hauser, Helwig

doi:10.1111/cgf.13405

Cited by 25 publications

(32 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Modeling attention can be a rich signal for inferring goals, intention and interest [Hor99a, HKPH03], and information about users' current and future attention can be useful for allocating computational resources [HKPH03] or for supporting data exploration [FH18]. For example, the system can perform pre‐computation or pre‐fetching based on its predictions.…”

Section: Discussionmentioning

confidence: 99%

“…They showed that off‐the‐shelf algorithms could successfully predict completion time and personality traits based on low‐level mouse clicks and moves [BOZ∗14]. Recent work by Fan et al used a convolution neural network to infer brush selection from a simple click and drag interaction design [FH18]. They demonstrated that their technique can quickly and accurately predict users' intended selections.…”

Section: Prior Work On Modeling and Predicting User Actionsmentioning

confidence: 99%

See 1 more Smart Citation

Follow The Clicks: Learning and Anticipating Mouse Interactions During Exploratory Data Analysis

Ottley

Garnett

Wan

2019

Computer Graphics Forum

View full text Add to dashboard Cite

The goal of visual analytics is to create a symbiosis between human and computer by leveraging their unique strengths. While this model has demonstrated immense success, we are yet to realize the full potential of such a human‐computer partnership. In a perfect collaborative mixed‐initiative system, the computer must possess skills for learning and anticipating the users' needs. Addressing this gap, we propose a framework for inferring attention from passive observations of the user's click, thereby allowing accurate predictions of future events. We demonstrate this technique with a crime map and found that users' clicks can appear in our prediction set 92% ‐ 97% of the time. Further analysis shows that we can achieve high prediction accuracy typically after three clicks. Altogether, we show that passive observations of interaction data can reveal valuable information that will allow the system to learn and anticipate future events.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Prior Work On Modeling and Predicting User Actionsmentioning

confidence: 99%

Follow The Clicks: Learning and Anticipating Mouse Interactions During Exploratory Data Analysis

Ottley

Garnett

Wan

2019

Computer Graphics Forum

View full text Add to dashboard Cite

show abstract

“…The community has also conducted several pieces of research on exploiting deep learning techniques to facilitate user interactions. Fan and Hauser [6] modeled user brushing in 2D scatterplot as an image, which can be handled by a convolutional neural network (CNN) to predict selected points. The method greatly improves selection accuracy, meanwhile preserves efficiency.…”

Section: Related Workmentioning

confidence: 99%

“…However, in our case, point clouds are distributed in 3D space. Thus CNNs (e.g., [6]) designed for 2D images are not feasible. Point clouds datasets exhibit great diversity, e.g., sparse vs dense, balanced vs imbalanced density.…”

Section: Related Workmentioning

confidence: 99%

LassoNet: Deep Lasso-Selection of 3D Point Clouds

Chen

Zeng

Yang

et al. 2020

IEEE Trans. Visual. Comput. Graphics

View full text Add to dashboard Cite

Fig. 1. LassoNet enables effective lasso-selection of 3D point clouds based on a latent mapping from viewpoint and lasso to target point clouds. LassoNet is particularly efficient for selecting multiple regions (insets 1, 2, 3) in a complex scene (left), since no viewpoint changing is required to select occluded points. Notice here the insets are viewed from different viewpoints.Abstract-Selection is a fundamental task in exploratory analysis and visualization of 3D point clouds. Prior researches on selection methods were developed mainly based on heuristics such as local point density, thus limiting their applicability in general data. Specific challenges root in the great variabilities implied by point clouds (e.g., dense vs. sparse), viewpoint (e.g., occluded vs. non-occluded), and lasso (e.g., small vs. large). In this work, we introduce LassoNet, a new deep neural network for lasso selection of 3D point clouds, attempting to learn a latent mapping from viewpoint and lasso to point cloud regions. To achieve this, we couple user-target points with viewpoint and lasso information through 3D coordinate transform and naive selection, and improve the method scalability via an intention filtering and farthest point sampling. A hierarchical network is trained using a dataset with over 30K lasso-selection records on two different point cloud data. We conduct a formal user study to compare LassoNet with two state-of-the-art lasso-selection methods. The evaluations confirm that our approach improves the selection effectiveness and efficiency across different combinations of 3D point clouds, viewpoints, and lasso selections. Project Website: https://lassonet.github.io

show abstract

“…Frey [Fre17] trained a neural network to determine the best progressive sampling strategy to calculate the similarity of spatio‐temporal data sets. In information visualization, Fan and Hauser [FH18] used CNNs to improve the manual brushing of points in scatterplots. Berger et al [BLL17] explored the use of generative adversary networks to analyze the role of transfer functions in the image synthesis process of direct volume rendering.…”

Section: Related Workmentioning

confidence: 99%

Robust Reference Frame Extraction from Unsteady 2D Vector Fields with Convolutional Neural Networks

Kim

Günther

2019

Computer Graphics Forum

View full text Add to dashboard Cite

Robust feature extraction is an integral part of scientific visualization. In unsteady vector field analysis, researchers recently directed their attention towards the computation of near‐steady reference frames for vortex extraction, which is a numerically challenging endeavor. In this paper, we utilize a convolutional neural network to combine two steps of the visualization pipeline in an end‐to‐end manner: the filtering and the feature extraction. We use neural networks for the extraction of a steady reference frame for a given unsteady 2D vector field. By conditioning the neural network to noisy inputs and resampling artifacts, we obtain numerically stabler results than existing optimization‐based approaches. Supervised deep learning typically requires a large amount of training data. Thus, our second contribution is the creation of a vector field benchmark data set, which is generally useful for any local deep learning‐based feature extraction. Based on Vatistas velocity profile, we formulate a parametric vector field mixture model that we parameterize based on numerically‐computed example vector fields in near‐steady reference frames. Given the parametric model, we can efficiently synthesize thousands of vector fields that serve as input to our deep learning architecture. The proposed network is evaluated on an unseen numerical fluid flow simulation.

show abstract

Fast and Accurate CNN‐based Brushing in Scatterplots

Cited by 25 publications

References 27 publications

Follow The Clicks: Learning and Anticipating Mouse Interactions During Exploratory Data Analysis

Follow The Clicks: Learning and Anticipating Mouse Interactions During Exploratory Data Analysis

LassoNet: Deep Lasso-Selection of 3D Point Clouds

Robust Reference Frame Extraction from Unsteady 2D Vector Fields with Convolutional Neural Networks

Contact Info

Product

Resources

About