About

Michael Gleicher is a David DeWitt Professor and Associate Chair for Undergraduate Programs in the Department of Computer Sciences at the University of Wisconsin, Madison. His research primarily focuses on Visual Computing, with current interests in robotics and data visualization, emphasizing how these technologies can be made useful for people. His work encompasses areas such as animation, virtual reality, multimedia, and visualization theory, including summarization, uncertainty, and effective visualization design. Gleicher is recognized as an ACM Fellow, a member of the IEEE Visualization Academy, and an IEEE Senior Member, and holds a concurrent position as an Amazon Design Scholar. Throughout his career, Gleicher has contributed to various themes in robotics and visualization, including shared autonomy for robotic inspection, communication of physical interactions between humans and robots, and the development of visualization methods that enhance understanding and decision-making. His research also explores novel sensors for robotics applications, such as Single Photon Avalanche Diode (SPAD) sensors, and investigates how emerging sensing technologies can be integrated into robotic systems. Additionally, he has worked on visualizing robot awareness, understanding human perception principles for visualization design, and improving image and video authoring techniques. Gleicher is actively involved in teaching, offering courses such as CS765 Visualization and CS559 Computer Graphics, and has a long history of teaching graduate and undergraduate classes in graphics and animation. He is also engaged in mentoring students and providing advice on research, projects, and academic development. His extensive publication record includes recent work in robotics, visualization, and related fields, reflecting his ongoing contributions to advancing knowledge in visual computing.

Research topics

• Computer Science
• Artificial Intelligence
• Information Retrieval
• Natural Language Processing
• Programming language
• Cartography
• Human–computer interaction
• Geography

Selected publications

• Augmenting a Large Language Model with a Combination of Text and Visual Data for Conversational Visualization of Global Geospatial Data

  arXiv (Cornell University) · 2025-01-16

  preprintOpen access

  We present a method for augmenting a Large Language Model (LLM) with a combination of text and visual data to enable accurate question answering in visualization of scientific data, making conversational visualization possible. LLMs struggle with tasks like visual data interaction, as they lack contextual visual information. We address this problem by merging a text description of a visualization and dataset with snapshots of the visualization. We extract their essential features into a structured text file, highly compact, yet descriptive enough to appropriately augment the LLM with contextual information, without any fine-tuning. This approach can be applied to any visualization that is already finally rendered, as long as it is associated with some textual description.

  PublisherOA PDFDOI

• Recovering Parametric Scenes from Very Few Time-of-Flight Pixels

  2025-10-19

  articleOpen access

  We aim to recover the geometry of 3D parametric scenes using very few depth measurements from low-cost, commercially available time-of-flight sensors. These sensors offer very low spatial resolution (i.e., a single pixel), but image a wide field-of-view per pixel and capture detailed time-of-flight data in the form of time-resolved photon counts. This time-of-flight data encodes rich scene information and thus enables recovery of simple scenes from sparse measurements. We investigate the feasibility of using a distributed set of few measurements (e.g., as few as 15 pixels) to recover the geometry of simple parametric scenes with a strong prior, such as estimating the 6D pose of a known object. To achieve this, we design a method that utilizes both feed-forward prediction to infer scene parameters, and differentiable rendering within an analysis-by-synthesis framework to refine the scene parameter estimate. We develop hardware prototypes and demonstrate that our method effectively recovers object pose given an untextured 3D model in both simulations and controlled real-world captures, and show promising initial results for other parametric scenes. We additionally conduct experiments to explore the limits and capabilities of our imaging solution.

  PublisherOA PDFDOI

• Development of the NBT-53 Texture Library for the Study of Texture Semantics

  Journal of Vision · 2025-07-15

  articleOpen access

  Texture is useful for representing categorical data in information visualizations, especially when color display capabilities are limited (He et al., 2024). Prior work on texture for visualization mostly focused on which dimensions of texture could represent data effectively, with little focus on texture semantics—the meaning people ascribe to textures. To study texture semantics (cf. color semantics; Schloss (2024), we need a library of visual textures that are relatively uniform, are perceptually discriminable, and span dimensions of texture perception. To develop such a library, we began with the Normalized Brodatz Texture (NBT) database, a standard set of texture images that have been normalized for lightness (Abdelmounaime & Dong-Chen, 2013). We aimed to subset the 112 textures to select textures that were (1) uniform over the image (important for future use in data visualizations) and (2) perceptually dissimilar. Toward these goals, we first asked participants to rate the uniformity of each texture, and we excluded textures that were, on average, below the neutral point of the rating scale. To assess the similarity of the remaining 62 textures, we asked a second group of participants to make triplet similarity judgements. Using data from over 11,000 trials across 59 participants, we estimated a 3-dimensional embedding of the textures that best explained human judgements (Sievert et al., 2023). Although similar embeddings exist for a subset of the original non-normalized Brodatz textures (Ravishankar Rao & Lohse, 1996), we reasoned that the dimensions could be different for the normalized images. The 3-dimensions of our embedding were: fine/course, hard/soft, and random/non-random (named using data from different participants). Finally, we used k-means clustering on this embedding to identify highly similar textures and selected the most uniform texture within each cluster. This approach yielded a final set of 53 textures, which we call the NBT-53 texture library.

  PublisherDOI

• Anytime Planning for End-Effector Trajectory Tracking

  ArXiv.org · 2025-02-05

  preprintOpen accessSenior author

  End-effector trajectory tracking algorithms find joint motions that drive robot manipulators to track reference trajectories. In practical scenarios, anytime algorithms are preferred for their ability to quickly generate initial motions and continuously refine them over time. In this paper, we present an algorithmic framework that adapts common graph-based trajectory tracking algorithms to be anytime and enhances their efficiency and effectiveness. Our key insight is to identify guide paths that approximately track the reference trajectory and strategically bias sampling toward the guide paths. We demonstrate the effectiveness of the proposed framework by restructuring two existing graph-based trajectory tracking algorithms and evaluating the updated algorithms in three experiments.

  PublisherOA PDFDOI

• Hierarchically Accelerated Coverage Path Planning for Redundant Manipulators

  2025-05-19

  articleSenior author

  Many robotic applications, such as sanding, polishing, wiping and sensor scanning, require a manipulator to dexterously cover a surface using its end-effector. In this paper, we provide an efficient and effective coverage path planning approach that leverages a manipulator's redundancy and task tolerances to minimize costs in joint space. We formulate the problem as a Generalized Traveling Salesman Problem and hierarchically streamline the graph size. Our strategy is to identify guide paths that roughly cover the surface and accelerate the computation by solving a sequence of smaller problems. We demonstrate the effectiveness of our method through a simulation experiment and an illustrative demonstration using a physical robot.

  PublisherDOI

• Efficient Detection of Objects Near a Robot Manipulator via Miniature Time-of-Flight Sensors

  ArXiv.org · 2025-09-19 · 1 citations

  preprintOpen accessSenior author

  We provide a method for detecting and localizing objects near a robot arm using arm-mounted miniature time-of-flight sensors. A key challenge when using arm-mounted sensors is differentiating between the robot itself and external objects in sensor measurements. To address this challenge, we propose a computationally lightweight method which utilizes the raw time-of-flight information captured by many off-the-shelf, low-resolution time-of-flight sensor. We build an empirical model of expected sensor measurements in the presence of the robot alone, and use this model at runtime to detect objects in proximity to the robot. In addition to avoiding robot self-detections in common sensor configurations, the proposed method enables extra flexibility in sensor placement, unlocking configurations which achieve more efficient coverage of a radius around the robot arm. Our method can detect small objects near the arm and localize the position of objects along the length of a robot link to reasonable precision. We evaluate the performance of the method with respect to object type, location, and ambient light level, and identify limiting factors on performance inherent in the measurement principle. The proposed method has potential applications in collision avoidance and in facilitating safe human-robot interaction.

  PublisherOA PDFDOI

• Hierarchically Accelerated Coverage Path Planning for Redundant Manipulators

  ArXiv.org · 2025-02-26

  preprintOpen accessSenior author

  Many robotic applications, such as sanding, polishing, wiping and sensor scanning, require a manipulator to dexterously cover a surface using its end-effector. In this paper, we provide an efficient and effective coverage path planning approach that leverages a manipulator's redundancy and task tolerances to minimize costs in joint space. We formulate the problem as a Generalized Traveling Salesman Problem and hierarchically streamline the graph size. Our strategy is to identify guide paths that roughly cover the surface and accelerate the computation by solving a sequence of smaller problems. We demonstrate the effectiveness of our method through a simulation experiment and an illustrative demonstration using a physical robot.

  PublisherOA PDFDOI

• Anytime Planning for End-Effector Trajectory Tracking

  IEEE Robotics and Automation Letters · 2025-02-11 · 2 citations

  articleSenior author

  End-effector trajectory tracking algorithms find joint motions that drive robot manipulators to track reference trajectories. In practical scenarios, anytime algorithms are preferred for their ability to quickly generate initial motions and continuously refine them over time. In this letter, we present an algorithmic framework that adapts common graph-based trajectory tracking algorithms to be anytime and enhances their efficiency and effectiveness. Our key insight is to identify guide paths that approximately track the reference trajectory and strategically bias sampling toward the guide paths. We demonstrate the effectiveness of the proposed framework by restructuring two existing graph-based trajectory tracking algorithms and evaluating the updated algorithms in three experiments.

  PublisherDOI

• Anchoring and Alignment: Data Factors in Part-to-Whole Visualization

  2025-11-01

  articleOpen accessSenior author

  We explore the effects of data and design considerations through the example case of part-to-whole data relationships. Standard part-to-whole representations like pie charts and stacked bar charts make the relationships of parts to the whole explicit. Value estimation in these charts benefits from two perceptual mechanisms: anchoring, where the value is close to a reference value with an easily recognized shape, and alignment where the beginning or end of the shape is aligned with a marker. In an online study, we explore how data and design factors such as value, position, and encoding together impact these effects in making estimations in part-to-whole charts. The results show how salient values and alignment to positions on a scale affect task performance. This demonstrates the need for informed visualization design based around how data properties and design factors affect perceptual mechanisms.

  PublisherOA PDFDOI

• Using a Distance Sensor to Detect Deviations in a Planar Surface

  IEEE Robotics and Automation Letters · 2024-08-19 · 5 citations

  articleSenior author

  We investigate methods for determining if a planar surface contains geometric deviations (e.g., protrusions, objects, divots, or cliffs) using only an instantaneous measurement from a miniature optical time-of-flight sensor. The key to our method is to utilize the entirety of information encoded in raw time-of-flight data captured by off-the-shelf distance sensors. We provide an analysis of the problem in which we identify the key ambiguity between geometry and surface photometrics. To overcome this challenging ambiguity, we fit a Gaussian mixture model to a small dataset of planar surface measurements. This model implicitly captures the expected geometry and distribution of photometrics of the planar surface and is used to identify measurements that are likely to contain deviations. We characterize our method on a variety of surfaces and planar deviations across a range of scenarios. We find that our method utilizing raw time-of-flight data outperforms baselines which use only derived distance estimates. We build an example application in which our method enables mobile robot obstacle and cliff avoidance over a wide field-of-view.

  PublisherDOI

Recent grants

• III: Small: Subset Selection for Summary Visualization

  NSF · $532k · 2020–2026

• NRI-Small: Perceptually Inspired Dynamics for Robot Arm Motion

  NSF · $800k · 2012–2018

• Collaborative Research: Retargetable Images and Video

  NSF · $406k · 2004–2011

• EAGER: Interacting with Machine Learning through Visual Comparison of Outcomes

  NSF · $186k · 2019–2022

• CGV: Medium: Collaborative Research: Visualizing Comparisons

  NSF · $599k · 2012–2019

Frequent coauthors

• Bilge Mutlu

  73 shared

• Daniel Rakita

  Yale University

  31 shared

• Michael Zinn

  University of Wisconsin–Madison

  29 shared

• Michael Hagenow

  22 shared

• Michael Ashikhmin

  20 shared

• Lucas Kovar

  University of Wisconsin–Madison

  20 shared

• Erik Reinhard

  University of Utah

  20 shared

• Robert G. Radwin

  University of Wisconsin–Madison

  20 shared

Education

• Ph.D., Computer Sciences

  University of Wisconsin, Madison

  1995

• M.S., Computer Sciences

  University of Wisconsin, Madison

  1991

• B.S., Computer Sciences

  University of Wisconsin, Madison

  1989

Awards & honors

• David DeWitt Professor of Computer Sciences
• ACM Fellow
• member of the IEEE Visualization Academy
• IEEE Senior Member
• Amazon Design Scholar