My main research revolves around the concept of Intelligence Augmentation, i.e. using information technology to supplement and support human thinking.
In particular, my interests focus on exploring how cutting-edge technologies can be combined to create hybrid, interactive systems able to augment human capabilities in collaborative settings. The final aim is to improve the way people communicate and interact with each other and with machines, building systems that will enhance how people currently perform their daily tasks and handle digital content. Sample applications include tools for interactive data exploration and scientific discovery, AI-assisted systems enabling the analysis of complex data, visual methods to interpret and automate non-trivial processes, and innovative interfaces to foster human productivity and collaboration.
My research interests lie at the intersection of several domains such as human computer interaction and user interfaces, data science and artificial intelligence, data visualization and computer graphics - with a particular passion for emerging technologies such as augmented and virtual reality.

Research statement, Marco Cavallo

Dataspace: A Reconfigurable Hybrid Reality Environment for Collaborative Information Analysis

Marco Cavallo, Mishal Dholakia, ..., Mark Podlaseck

Immersive environments have gradually become standard for visualizing and analyzing large or complex datasets that would otherwise be cumbersome, if not impossible, to explore through smaller scale computing devices. However, this type of workspace often proves to possess limitations in terms of interaction, flexibility, cost and scalability. In this paper we introduce a novel immersive environment called Dataspace, which features a new combination of heterogeneous technologies and methods of interaction towards creating a better team workspace. Dataspace provides 15 high-resolution displays that can be dynamically reconfigured in space through robotic arms, a central table where information can be projected, and a unique integration with augmented reality (AR) and virtual reality (VR) headsets and other mobile devices. In particular, we contribute novel interaction methodologies to couple the physical environment with AR and VR technologies, enabling visualization of complex types of data and mitigating the scalability issues of existing immersive environments. We demonstrate through four use cases how this environment can be effectively used across different domains and reconfigured based on user requirements. Finally, we compare Dataspace with existing technologies, summarizing the trade-offs that should be considered when attempting to build better collaborative workspaces for the future.

Clustrophile 2: Guided Visual Clustering Analysis

Marco Cavallo, Cagatay Demiralp

Data clustering is a common unsupervised learning method frequently used in exploratory data analysis. However, identifying relevant structures in unlabeled, high-dimensional data is nontrivial, requiring iterative experimentation with clustering parameters as well as data features and instances. The space of possible clusterings for a typical dataset is vast, and navigating in this vast space is also challenging. The absence of ground-truth labels makes it impossible to define an optimal solution, thus requiring user judgment to establish what can be considered a satisfiable clustering result. Data scientists need adequate interactive tools to effectively explore and navigate the large space of clusterings so as to improve the effectiveness of exploratory clustering analysis.
We introduce Clustrophile 2, a new interactive tool for guided clustering analysis. Clustrophile 2 guides users in clustering-based exploratory analysis, adapts user feedback to improve user guidance, facilitates the interpretation of clusters, and helps quickly reason about differences between clusterings. To this end, Clustrophile 2 contributes a novel feature, the clustering tour, to help users choose clustering parameters and assess the quality of different clustering results in relation to current analysis goals and user expectations.
We evaluate Clustrophile 2 through a user study with 12 data scientists, who used our tool to explore and interpret sub-cohorts in a dataset of Parkinsons disease patients. Results suggest that Clustrophile 2 improves the speed and effectiveness of exploratory clustering analysis for both experts and non-experts.

A Visual Interaction Framework for Dimensionality Reduction Based Data Exploration

Marco Cavallo, Cagatay Demiralp

Dimensionality reduction is a common method for analyzing and visualizing high-dimensional data. However, reasoning dynamically about the results of a dimensionality reduction is difficult. Dimensionality-reduction algorithms use complex optimizations to reduce the number of dimensions of a dataset, but these new dimensions often lack a clear relation to the initial data dimensions, thus making them difficult to interpret. Here we propose a visual interaction framework to improve dimensionality-reduction based exploratory data analysis. We introduce two interaction techniques, forward projection and backward projection, for dynamically reasoning about dimensionally reduced data. We also contribute two visualization techniques, prolines and feasibility maps, to facilitate the effective use of the proposed interactions. We apply our framework to PCA and autoencoder-based dimensionality reductions. Through data-exploration examples, we demonstrate how our visual interactions can improve the use of dimensionality reduction in exploratory data analysis.

Track Xplorer: A System for Visual Analysis of Sensor-based Motor Activity Predictions

Marco Cavallo, Cagatay Demiralp

Detecting motor activities from sensor datasets is becoming increasingly common in a wide range of applications with the rapid commoditization of wearable sensors. To detect activities, data scientists iteratively experiment with different classifiers before deciding on a single model. Evaluating, comparing, and reasoning about prediction results of alternative classifiers is a crucial step in the process of iterative model development. However, standard aggregate performance metrics (such as accuracy score) and textual display of individual event sequences have limited granularity and scalability to effectively perform this critical step. To ameliorate these limitations, we introduce Track Xplorer, an interactive visualization system to query, analyze and compare the classification output of activity detection in multi-sensor data. Track Xplorer visualizes the results of different classifiers as well as the ground truth labels and the video of activities as temporally-aligned linear tracks. Through coordinated track visualizations, Track Xplorer enables users to interactively explore and compare the results of different classifiers, assess their accuracy with respect to the ground truth labels and video. Users can brush arbitrary regions of any classifier track, zoom in and out with ease, and playback the corresponding video segment to contextualize the performance of the classifier within the selected region. Track Xplorer also contributes an algebra over track representations to filter, compose, and compare classification outputs, enabling users to effectively reason about the performance of classifiers. We demonstrate how our tool helps data scientists debug misclassifications and improve the prediction performance in developing activity classifiers for real-world, multi-sensor data gathered from Parkinson's patients.

Decomposition of complex movements into primitives for Parkinson's disease assessment

Eleftheria Pissadaki, ..., Marco Cavallo et al.

Recent advances in technology present an important opportunity in medicine to augment episodic, expert-based observations of patients' disease signs, obtained in the clinic, with continuous and sensitive measures using wearable and ambient sensors. In Parkinson's disease (PD), such technology-based objective measures have shown exciting potential for passively monitoring disease signs, their fluctuation, and their progression. We are developing a system to passively and continuously capture data from people with PD in their daily lives, and provide a real-time estimate of their motor functions, that is analogous to scores obtained during Part III of the humanadministered Movement Disorder Society's Unified Parkinson's Disease assessment (MDS-UPDRS3). Our hypothesis is that complex human movements can be decomposed into movement primitives related to the performance of the MDS-UPDRS3 motor assessment. Toward this hypothesis, we developed a system for integrating and analyzing multiple streams of sensor data collected from volunteers executing the tasks based on the MDS-UPDRS3. In this paper, we show how we can leverage the data collected from MDS-UPDRS3 tasks to develop machine learning models that can identify movement primitives in activities of daily living.

Exploring Dimensionality Reductions with Forward and Backward Projections

Marco Cavallo, Cagatay Demiralp

Dimensionality reduction is a common method for analyzing and visualizing high-dimensional data across domains. Dimensionality-reduction algorithms involve complex optimizations and the reduced dimensions computed by these algorithms generally lack clear relation to the initial data dimensions. Therefore, interpreting and reasoning about dimensionality reductions can be difficult. In this work, we introduce two interaction techniques, forward projection and backward projection, for reasoning dynamically about scatter plots of dimensionally reduced data. We also contribute two related visualization techniques, prolines and feasibility map to facilitate and enrich the effective use of the proposed interactions, which we integrate in a new tool called Praxis. To evaluate our techniques, we first analyze their time and accuracy performance across varying sample and dimension sizes. We then conduct a user study in which twelve data scientists use Praxis so as to assess the usefulness of the techniques in performing exploratory data analysis tasks. Results suggest that our visual interactions are intuitive and effective for exploring dimensionality reductions and generating hypotheses about the underlying data.

CAVE-AR: A VR Authoring System to Interactively Design, Simulate, and Debug Multi-user AR Experiences

Marco Cavallo, Angus Forbes

In this paper we propose CAVE-AR, a novel virtual reality (VR) system for authoring custom augmented reality (AR) experiences and interacting with participating users. We introduce an innovative technique to integrate different representations of the world, mixing geographical information, architectural features, and sensor data, allowing us to understand precisely how users are behaving within the AR experience. By taking advantage of this technique to "mix realities", our VR application provides the designer with tools to create and modify a AR application, even while other people are in the midst of using it. Our VR application further lets the designer track how users are behaving, preview what they are currently seeing, and interact with them through different channels. This enables new possibilities which range from simple debugging and testing to more complex forms of centralized task control, such as placing a virtual avatar in the AR experience to guide a user. In addition to describing details of how we create effective representations of the real-world for enhanced AR experiences and our novel interaction modalities, we introduce two use cases demonstrating the potential of our approach. The first is an AR experience that enables users to discover historical information during an urban tour along the Chicago Riverwalk; the second is a novel scavenger hunt that places virtual objects within a realworld environment to facilitate solving complex multi-user puzzles. In both cases, the ability to develop and test the AR experience remotely greatly enhanced the design process and the novel interaction techniques greatly enhanced overall user experience.

Riverwalk: Incorporating Historical Photographs in Public Outdoor Augmented Reality Experiences

Marco Cavallo, Angus Forbes, Geoffrey Alan Rhodes

We introduce a user-centered Augmented Reality (AR) approach for publishing 2D media archives as interactive content. We discuss the relevant technical considerations for developing an effective application for public outdoor AR experiences that leverage context-specific elements in a challenging, real-world environment. Specifically, we show how a classical marker-less approach can be combined with mobile sensors and geospatial information in order apply our knowledge of the surroundings to the experience itself. Our contributions provide the enabling technology for Chicago 0,0 Riverwalk, a novel app-based AR experience that superimposes historical imagery onto matching views in downtown Chicago, Illinois along an open, pedestrian waterfront located on the bank of the Chicago River. Historical photographs of sites along the river are superimposed onto buildings, bridges, and other architectural features through image-based AR tracking, providing a striking experience of the city's history as rooted in extant locations along the river.

RehabJim: A Third Person Approach To Virtual Reality Biomechanical Rehabilitation

Marco Cavallo, Andrea Rottigni, Elizabeta Marai, James Patton

RehabJim is a Unity3D application developed on behalf of researchers at a major Rehabilitation Institute in order to explore the opportunities that Virtual Reality may offer to biomechanical and neurological rehabilitation. In particular, our work focuses on arm actions peformed while standing by patients post-stroke. We leverage a 3D immersive environment, augmented with Kinect interaction. Our approach follows a third person perspective of the patient's body, and employs two cartoon-style avatars - the patient and their therapist. The user is asked to reach to computer-generated virtual objects with her hands. Our evaluation with users and a domain expert shows that this type of environment leads to an engaging, enjoyable experience that can encourage patients to perform a wide variety of whole-body motions.

DigitalQuest: A Mixed Reality Approach to Scavenger Hunts

Marco Cavallo, Angus Forbes

This paper presents a novel approach for the design of creative location-based mixed reality applications. We introduce a framework called DigitalQuest that simplifies adding geolocated virtual content on top of real-world camera input. Unlike previous work, which relies solely on markers or image pattern recognition, we define a "mirror world" that facilitates interactive mixed reality. DigitalQuest consists of an editor that allows users to easily add their own content as desired and a mobile application that loads content from a server based on the location of the device. Each piece of virtual content can be organized through the editor so that it appears only in certain circumstances, allowing a designer to determine when and where a virtual object is attached to a real-world location. We have used our editor to create a series of futuristic scavenger hunts in which participating teams must solve puzzles in order to access new virtual context appearing in a mixed reality environment via a mobile phone application. In this paper, we introduce our editor and present an example scavenger hunt game, Morimondo, that was built using it. Specifically, we describe our technique to utilize camera and motion sensors on the mobile phone to enable an appropriate level of user engagement within this game. We are able to obtain realistic augmentations with accurate positioning by leveraging sensor fusion and through the use of filters that compensate for sensor noise, using image processing only for error correction or in special situations. The initial success of this project leads us to believe that DigitalQuest could be used to design a wide range of creative multi-user mixed reality applications.

3D City Reconstruction From Google Street View

Marco Cavallo

Despite laser scan 3D point cloud acquisition has greatly improved over the next few years, the process of creating 3D large scale city models is still quite expensive and not straightforward. At the same time, nowadays services such as Google Street View provide a vast amount of geo-registered panoramic imagery, guaranteeing a decent resolution for dense locations at zero cost. Our idea is indeed to leverage this free information provided by Google Street View in order to obtain a cheap and automatizable 3D recontruction of an urban area, by extracting the depth information related to the great number of panoramic images available online.

Website obviously created by me :P.. Some pages are still under construction, my apologies for that.