Goki Muramoto
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Goki Muramoto has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Masahiko Inami
University of Tokyo, Tokyo, Japan and Research Center for Advanced Science and Technology, Tokyo, Japan
4 shared publications - Hiroto Saito
The University of Tokyo, Tokyo, Japan
3 shared publications - Hirosuke Asahi
The University of Tokyo, Tokyo, Japan
2 shared publications - Sohei Wakisaka
Keio University Graduate School of Media Design, Yokohama, Japan
2 shared publications - Yuri Yasui
The University of British Columbia, Vancouver, British Columbia, Canada
2 shared publications - Jumpei Yamashita
Information Physics and Computing, the University of Tokyo, Tokyo, Japan
1 shared publication
Latest Publications
- Open Access
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1867-1882
Semantic See-through Goggles: Wearing Linguistic Virtual Reality in (Artificial) Intelligence
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1867-1882
Open AccessSemantic See-through Goggles: Wearing Linguistic Virtual Reality in (Artificial) Intelligence
- Open Access
- 16 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 202-212
Shared Gaze on AI Node Maps: Mitigating Empathy Fog inAI-Mediated Dyadic Discussion
Recent AI-Mediated Communication (AI-MC) systems use large language models to summarize, rephrase, and structure dialogue into shared AI node maps, externalizing the evolving argument structure and reducing cognitive load. However, such artifacts can create an “Empathy Fog,” where attention shifts from the human partner to the interface and social presence weakens. We examine whether overlaying Shared Gaze Visualization (SGV) on AI node maps can mitigate this trade-off. We conducted an exploratory laboratory study with 16 participants (8 dyads) solving Hidden Profile tasks under three within-subject conditions (face-to-face, AI node map only, AI node map with SGV). Compared to the face-to-face baseline, the map-only condition was associated with a shift in the locus of interaction toward the tool and with an absence of creative hypotheses. Adding SGV increased perceived surveillance but did not induce discomfort, and instead coincided with more creative hypotheses and a stronger sense of “talking to a person.” We derive design implications for balancing AI-generated structure with shared gaze in AI-mediated dyadic discussion.
- 16 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 202-212
Open AccessShared Gaze on AI Node Maps: Mitigating Empathy Fog inAI-Mediated Dyadic Discussion
Recent AI-Mediated Communication (AI-MC) systems use large language models to summarize, rephrase, and structure dialogue into shared AI node maps, externalizing the evolving argument structure and reducing cognitive load. However, such artifacts can create an “Empathy Fog,” where attention shifts from the human partner to the interface and social presence weakens. We examine whether overlaying Shared Gaze Visualization (SGV) on AI node maps can mitigate this trade-off. We conducted an exploratory laboratory study with 16 participants (8 dyads) solving Hidden Profile tasks under three within-subject conditions (face-to-face, AI node map only, AI node map with SGV). Compared to the face-to-face baseline, the map-only condition was associated with a shift in the locus of interaction toward the tool and with an absence of creative hypotheses. Adding SGV increased perceived surveillance but did not induce discomfort, and instead coincided with more creative hypotheses and a stronger sense of “talking to a person.” We derive design implications for balancing AI-generated structure with shared gaze in AI-mediated dyadic discussion.
- Open Access
- 3 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 1
Semantic See-through Goggles
Semantic See-through Goggles are a kind of glasses through which the view is once fully verbalized and re-imaged. Participants wear original goggles with a camera and HMD. First, a single image from the camera is converted by the AI into a single line of text. Next, an image is generated from that line of text, also by the AI. Finally, this image is presented on the display in both eyes. Both input and output are concrete images, but only the information contained in one line of text is stored, all the rest is creative. The view, in which much of the spatial positioning and material has been reworked, however, does indeed convey the meaning of the scene in front of it. These processes take place in real time, allowing the user to perceive and act on the real physical world through this redrawn view. Every view in the goggles is a 'stereotype' of some kind, which is constructed on the basis of the specific cultural knowledge learned by the model. (For example, the view of a person looking at a 'longhaired Asian man', oscillates between a standard Anglo-Saxon, standard-bodied man and an equally Anglo-Saxon, standard-bodied woman.) This experience provides a subjective understanding of what it is like to divide the task of perceiving the world with an AI, while at the same time reminds us of the basic and inescapable fact of human perception that 'We see the world semantically'.
- 3 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 1
Open AccessSemantic See-through Goggles
Semantic See-through Goggles are a kind of glasses through which the view is once fully verbalized and re-imaged. Participants wear original goggles with a camera and HMD. First, a single image from the camera is converted by the AI into a single line of text. Next, an image is generated from that line of text, also by the AI. Finally, this image is presented on the display in both eyes. Both input and output are concrete images, but only the information contained in one line of text is stored, all the rest is creative. The view, in which much of the spatial positioning and material has been reworked, however, does indeed convey the meaning of the scene in front of it. These processes take place in real time, allowing the user to perceive and act on the real physical world through this redrawn view. Every view in the goggles is a 'stereotype' of some kind, which is constructed on the basis of the specific cultural knowledge learned by the model. (For example, the view of a person looking at a 'longhaired Asian man', oscillates between a standard Anglo-Saxon, standard-bodied man and an equally Anglo-Saxon, standard-bodied woman.) This experience provides a subjective understanding of what it is like to divide the task of perceiving the world with an AI, while at the same time reminds us of the basic and inescapable fact of human perception that 'We see the world semantically'.
- Open Access
- 13 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-2
Demonstrating real-time slow-motion experience through parallel video presentation
Real-time and slow-motion are basically incompatible properties in the technologies of recording and playback. We propose a framework to realize the coexistence of both using parallel video presentation based on the integrative capacity and the temporal nature of cognition. In this framework, camera input is divided into short time durations, distributed into layered timelines, and then stretched out. And then, all layers are composited and presented in a display simultaneously. In our demonstration, users can experience their surroundings through see-through goggles or digital mirrors in which this framework is embedded. Under certain conditions, the perception of the layers is integrated into the cognitive process, resulting in the coexistence of real-time and slow-motion. This research suggests a new field of “Temporal Editing to Humans.”
- 13 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-2
Open AccessDemonstrating real-time slow-motion experience through parallel video presentation
Real-time and slow-motion are basically incompatible properties in the technologies of recording and playback. We propose a framework to realize the coexistence of both using parallel video presentation based on the integrative capacity and the temporal nature of cognition. In this framework, camera input is divided into short time durations, distributed into layered timelines, and then stretched out. And then, all layers are composited and presented in a display simultaneously. In our demonstration, users can experience their surroundings through see-through goggles or digital mirrors in which this framework is embedded. Under certain conditions, the perception of the layers is integrated into the cognitive process, resulting in the coexistence of real-time and slow-motion. This research suggests a new field of “Temporal Editing to Humans.”
- Open Access
- 4 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 91-101
Real-time Slow-motion : A Framework for Slow-motion Without Deviating from Real-time
Real-time and slow-motion are incompatible. This is a fundamental incompatibility based on the relationship between recording and playback. We propose Real-time Slow-motion, a framework that allows for the coexistence of both. In a see-through system with a pair of a camera and head-mounted display (HMD), this framework provides temporal editing between camera and display within the short width of the subjective present (i.e. now): the camera inputs, divided into small durations, are alternately distributed on two timelines, then stretched twice on the timelines, and these layers are assigned to the left and right eye displays respectively. The hypothesis is that, under certain conditions, the brain integrates images from both eyes, resulting in a slow-motion experience that does not deviate from real-time (i.e., users experience real-time events with slow-motion mode). This paper describes the implementation and the psychological evaluation experiments and interviews about the realized experience. The framework of Real-time Slow-motion is then generalized, and its applications and limitations are discussed. This research suggests a new field of “Temporal Editing to Humans.”
- 4 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 91-101
Open AccessReal-time Slow-motion : A Framework for Slow-motion Without Deviating from Real-time
Real-time and slow-motion are incompatible. This is a fundamental incompatibility based on the relationship between recording and playback. We propose Real-time Slow-motion, a framework that allows for the coexistence of both. In a see-through system with a pair of a camera and head-mounted display (HMD), this framework provides temporal editing between camera and display within the short width of the subjective present (i.e. now): the camera inputs, divided into small durations, are alternately distributed on two timelines, then stretched twice on the timelines, and these layers are assigned to the left and right eye displays respectively. The hypothesis is that, under certain conditions, the brain integrates images from both eyes, resulting in a slow-motion experience that does not deviate from real-time (i.e., users experience real-time events with slow-motion mode). This paper describes the implementation and the psychological evaluation experiments and interviews about the realized experience. The framework of Real-time Slow-motion is then generalized, and its applications and limitations are discussed. This research suggests a new field of “Temporal Editing to Humans.”
- Open Access
- 25 July 2022
- Published by Association for Computing Machinery (ACM)
- p. 1
Imagraph
Imagraph is a medium that arbitrates between two primordial "attitudes": projecting the images and closing the senses. Participants lie on their backs in a box-shaped bed and shut their eyes. Two pieces of LED display board are suspended above their heads, with optical fibers from each pixel extending to the surface of their eyelids. The video prepared in advance is unilaterally played after the spectral compensation against the "blood-red" unique to their own flesh. There, the eyelid becomes the medium for the very object it is trying to block off. The intended colors and their placements are sent, but this in no way implies the triumph of projection over closing. The ritualistic posture forms a distinctive video place filled with the "tone of the unconscious." Just as eye closure deprives the participant of the rejection capability, projection as the arrayed commands also loses its authority. Participants unintentionally derive movements and colors that are not actually presented to them and blend the internally sprung image into the video irreversibly. Both projecting and closing are dysfunctional, and a singular relationship is established between the "projector" and the "projectee." What is the freedom each pole has, and where does the "image" inhabit?
- 25 July 2022
- Published by Association for Computing Machinery (ACM)
- p. 1
Open AccessImagraph
Imagraph is a medium that arbitrates between two primordial "attitudes": projecting the images and closing the senses. Participants lie on their backs in a box-shaped bed and shut their eyes. Two pieces of LED display board are suspended above their heads, with optical fibers from each pixel extending to the surface of their eyelids. The video prepared in advance is unilaterally played after the spectral compensation against the "blood-red" unique to their own flesh. There, the eyelid becomes the medium for the very object it is trying to block off. The intended colors and their placements are sent, but this in no way implies the triumph of projection over closing. The ritualistic posture forms a distinctive video place filled with the "tone of the unconscious." Just as eye closure deprives the participant of the rejection capability, projection as the arrayed commands also loses its authority. Participants unintentionally derive movements and colors that are not actually presented to them and blend the internally sprung image into the video irreversibly. Both projecting and closing are dysfunctional, and a singular relationship is established between the "projector" and the "projectee." What is the freedom each pole has, and where does the "image" inhabit?