2016 Fiscal Year Final Research Report
Fusion of Object Recognition and Material Recognition for Smarter Image Recognition
| Project/Area Number |
26540078
|
|---|
| Research Category |
Grant-in-Aid for Challenging Exploratory Research
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Perceptual information processing
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
|
|---|
| Research Collaborator |
AIZAWA Kiyoharu
|
|---|
| Project Period (FY) |
2014-04-01 – 2017-03-31
|
| Keywords | 物体認識 / 物体検出 / 確信度 / 文脈 / 特徴抽出 |
|---|
| Outline of Final Research Achievements |
There are three major contributions in our research. The first contribution was the confidence analysis for multi-class object recognition using the intermediate values from machine learning algorithms. By using the confidence, it has been made possible to improve the recognition accuracy. The second contribution was the methods to automatically find optimal parameter settings for convolutional neural networks (CNNs) by using an evolutionary algorithm called particle swarm optimization (PSO). We have also developed two candidate pruning algorithms for efficient evolutionary process. The third contribution was taking contextual information into consideration such as the co-occurrence of objects and the location of objects in object detection. We developed candidate pruning and object rescoring methods that leverage contextual information and that can improve the state-of-the-art CNN-based object detection methods such as Fast R-CNN and Faster R-CNN.
|
| Free Research Field |
画像処理、パターン認識、機械学習
|
