| Research Topics |
In addition to the research described below, we are studying a variety of topics. |
| ・Visualizing positive/negative contribution related to decision with DNN |
Visualizing spatial contribution for the decision of classification of diseases, which is one of the application of this method.
|
|
An example of applying tomogram of the fundus of an eye |
| |
・Implicit gaze induction by virtual shadow casting |
A trial to implicit gaze induction to a real object.
|
|
Virtual shadow casting and gaze heatmap |
| |
| ・Recognition of Human-Human interaction and Human-Object interaction |
Recognizing semantics of human-human interaction and human-object interaction with NN
|
|
Processing diagram for human-human interaction |
|
Processing diagram for human-onject interaction |
| |
| ・Fault detection for industrial products |
We are developing systems for fault detection of industrial products
|
| ・Protozoa detection and identification in micrograph |
DNN detects and identifies the species of protozoa in micrograph.
|
|
An example of detection and identification (Balantidium) |
| |
| ・Video processing for traffic monitoring systems |
Vehicle density estimation with DNN incorporated with "backward connection".
|
|
An example of estimation（Center: without backward connection, Right: with backward propagation) |
| |
| ・Headlight detection for night-time traffic surveillance |
Proposed method spparates headlight from its reflection for more accurate traffic counting based on spatiotemporal information.
|
|
An example of detection |
| |
| ・Subtype classification of AML with detection of white bolld cells |
Subtype of AML is classified after detecting the region of white blood cell from micrograph of humann blood
|
|
An example of white bolld cell detection based on its color and gradient |
| |
| ・Intuitive and easy to use interface for migrating paper-printed information |
Easy and intuitive operation with mobile device such as tablet or smartphone with camera enables to migrate papre-printed information into electrical data.
|
|
A snapshot of the interface for wasy data migration |
| |
| ・Inflextion point detection for TGGE |
Detecting inflextion points for DNA analysis with TGGE.
|
| ・Content-Based Retrieval based on Film Grammar |
Searching chronological contents, such as movies, drama, etc., is often a time-consuming task,
since it is not satisfactory to see just some screen shots to catch the whole story.
Summarizing such contents is one solution to reduce the time burden in browsing, and to allow
people to grasp the contents rapidly. Most of the video summarization methods previously proposed
extract 'conspicuous' shots from audio/visual point of view, and combine them together
to create a summary. These methods have the disadvantage of disregarding contextual dependency
between shots or scenes. We propose a method of summary generation based on 'film grammar',
which maintains the dependency between shots or scenes. Experimental results show this method
provides a viewer better summaries, for better comprehension of the context.
|
|
Detecting expressions of affective information and visualizing it with graphical representations. |
| |
|
Detecting various types of camera work from video tomography
for the evaluation of affective information. |
| |
| ・Multimedia data retrieval |
Video data is often acompanied with audio data. We are studying methods to detect implicit information
that is detected by the analysis of mutual relationship between audio and video data.
|
| ・Visual information parsing and interaction based on gaze detection |
Human behavior analysis for implicit preference detection and information recommendation based on gaze detection.
|
|
A system to recommend preferred information
based on imolicit interest detection based on gaze detection |
| |
| ・Video/Photo shooting assistance for better represenatation of affective information |
Three are several points of view to understand "well organized picture/video".
We focus on composition and depth of field in order to assist non-professional user to adjust or decide
better composition or the depth of field for individual occasion.
|
|
Camera system enabling to assist the decision of composition |
|
Camera system to assist better adjustment of depth of field |
| |
|
Video camera which enables to guide a user to preform proper camerawork for better representation of affective informatoin
|
| |
