Ian M. Thomas
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Ian M. Thomas has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Alberto Salleo
Department of Materials Science and Engineering Stanford University Stanford California USA
1 shared publication - Danielle Albers Szafir
University of North Carolina-Chapel Hill, Chapel Hill, North Carolina, USA
1 shared publication David W. Camp
Lawrence Livermore National Lab. (USA)
1 shared publicationFrancois Y. Genin
Lawrence Livermore National Lab. (USA)
1 shared publicationIan D. Hutcheon
Lawrence Livermore National Lab. (USA)
1 shared publicationJ. M. Yoshiyama
Lawrence Livermore National Lab. (USA)
1 shared publication
Latest Publications
- Open Access
- 18 March 2024
- Published by Association for Computing Machinery (ACM)
- p. 772-786
Assessing User Trust in Active Learning Systems: Insights from Query Policy and Uncertainty Visualization
Active learning (AL) systems have become increasingly popular for various applications in machine learning (ML), including medical imaging, environmental monitoring, and geospatial analysis. These systems rely on inputs dynamically queried from people to enhance classification. Ensuring appropriate analyst trust in these systems presents a significant obstacle as analyst over- or underreliance may adversely affect a given application. Common AL strategies enhance classification models by asking analysts to provide labels for data points with the highest degree of uncertainty. However, such model-centric policies do not consider potential priming effects on the analyst and how they will affect people’s trust in the system post-training. We present an empirical study assessing how AL query policies and visualizations that enhance transparency in a classifier’s decisions influence trust in automated image classifiers. We found that query policy may significantly influence an analyst’s perception of system capabilities, while the level of visual transparency into classifier certainty may influence an analyst’s ability to perform a classification task. Our study informs the design of interactive labeling systems to help mitigate the effects of overreliance and calibrate appropriate trust in automated systems.
- 18 March 2024
- Published by Association for Computing Machinery (ACM)
- p. 772-786
Open AccessAssessing User Trust in Active Learning Systems: Insights from Query Policy and Uncertainty Visualization
Active learning (AL) systems have become increasingly popular for various applications in machine learning (ML), including medical imaging, environmental monitoring, and geospatial analysis. These systems rely on inputs dynamically queried from people to enhance classification. Ensuring appropriate analyst trust in these systems presents a significant obstacle as analyst over- or underreliance may adversely affect a given application. Common AL strategies enhance classification models by asking analysts to provide labels for data points with the highest degree of uncertainty. However, such model-centric policies do not consider potential priming effects on the analyst and how they will affect people’s trust in the system post-training. We present an empirical study assessing how AL query policies and visualizations that enhance transparency in a classifier’s decisions influence trust in automated image classifiers. We found that query policy may significantly influence an analyst’s perception of system capabilities, while the level of visual transparency into classifier certainty may influence an analyst’s ability to perform a classification task. Our study informs the design of interactive labeling systems to help mitigate the effects of overreliance and calibrate appropriate trust in automated systems.
- 20 April 1998
- Published by SPIE-Intl Soc Optical Eng
- p. 331-340
Effects of polishing, etching, cleaving, and water leaching on the UV laser damage of fused silica
A damage morphology study was performed with a 355 nm, 8-ns Nd:YAG laser on synthetic UV-grade fused silica to determine the effects of post-polish chemical etching on laser-induced damage, compare damage morphologies of cleaved and polished surfaces, and understand the effects of the hydrolyzed surface layer and water-crack interactions. The samples were polished, then chemically etched in a buffered HF solution to remove 45, 90, 135, and 180 nm of surface material. Another set of sample was cleaved and soaked in boiling distilled water for 1 second and 1 hour. All the samples were irradiated at damaging fluences and characterized by Normarski optical microscopy and scanning electron microscope. Damage was initiated at micro-pits on both input and output surface of the polished fused silica sample. At higher fluences, the micro-pits generated cracks on the surface. Laser damage of the etched fused silica surface shoed that the real density of micro-pits decreased with etched thickness. SIMS analysis of the polished surface showed significant trace contamination levels within a 50 nm surface layer. Micro-pits formation also appeared after irradiating cleaved fused silica surfaces at damaging fluences. Linear damage tracks corresponding cleaving cracks were often observed on cleaved surfaces. Soaking cleaved samples in water produced wide laser damage tracks.
- 20 April 1998
- Published by SPIE-Intl Soc Optical Eng
- p. 331-340
Effects of polishing, etching, cleaving, and water leaching on the UV laser damage of fused silica
A damage morphology study was performed with a 355 nm, 8-ns Nd:YAG laser on synthetic UV-grade fused silica to determine the effects of post-polish chemical etching on laser-induced damage, compare damage morphologies of cleaved and polished surfaces, and understand the effects of the hydrolyzed surface layer and water-crack interactions. The samples were polished, then chemically etched in a buffered HF solution to remove 45, 90, 135, and 180 nm of surface material. Another set of sample was cleaved and soaked in boiling distilled water for 1 second and 1 hour. All the samples were irradiated at damaging fluences and characterized by Normarski optical microscopy and scanning electron microscope. Damage was initiated at micro-pits on both input and output surface of the polished fused silica sample. At higher fluences, the micro-pits generated cracks on the surface. Laser damage of the etched fused silica surface shoed that the real density of micro-pits decreased with etched thickness. SIMS analysis of the polished surface showed significant trace contamination levels within a 50 nm surface layer. Micro-pits formation also appeared after irradiating cleaved fused silica surfaces at damaging fluences. Linear damage tracks corresponding cleaving cracks were often observed on cleaved surfaces. Soaking cleaved samples in water produced wide laser damage tracks.
- Open Access
- 7 September 1994
- Published by SPIE-Intl Soc Optical Eng
- Vol. 2262, 54-59
Future trends in optical coatings for high-power laser applications
Inertial Confinement Fusion (ICF) research has historically been a driver in the development of high performance, high damage threshold optical coatings. This is particularly the case now as the ICF community develops plans for a proposed 1.8 mega-joule solid state (Nd+3-phosphate glass) laser system. The new system, the National Ignition Facility, is possible in part due to advances in optical coatings technology including the laser-conditioning of multilayer dielectrics and broadened applications for room-temperature deposited coatings. Sol-gel AR coatings are the standard for large, high-power laser optics and sol-gel HR coatings are being developed. For mirror and polarizer coatings, e-beam-deposited dielectrics continue to provide the highest damage threshold coatings, but their laser damage thresholds and optical performance are limited by micrometers -scale defects and poor control over layer thickness, respectively. More energetic deposition techniques such as IAD and IBS, now popular in the commercial market, offer both advantages and disadvantages in this high-damage- threshold coatings market.
- 7 September 1994
- Published by SPIE-Intl Soc Optical Eng
- Vol. 2262, 54-59
Open AccessFuture trends in optical coatings for high-power laser applications
Inertial Confinement Fusion (ICF) research has historically been a driver in the development of high performance, high damage threshold optical coatings. This is particularly the case now as the ICF community develops plans for a proposed 1.8 mega-joule solid state (Nd+3-phosphate glass) laser system. The new system, the National Ignition Facility, is possible in part due to advances in optical coatings technology including the laser-conditioning of multilayer dielectrics and broadened applications for room-temperature deposited coatings. Sol-gel AR coatings are the standard for large, high-power laser optics and sol-gel HR coatings are being developed. For mirror and polarizer coatings, e-beam-deposited dielectrics continue to provide the highest damage threshold coatings, but their laser damage thresholds and optical performance are limited by micrometers -scale defects and poor control over layer thickness, respectively. More energetic deposition techniques such as IAD and IBS, now popular in the commercial market, offer both advantages and disadvantages in this high-damage- threshold coatings market.