Automatic Incremental Model Learning for Scene Interpretation
نویسنده
چکیده
In this paper, we investigate automatic model learning for the interpretation of complex scenes with structured objects. We present a learning, interpretation, and evaluation cycle for processing such scenes. By including learning and interpretation in one framework, an evaluation and feedback learning is enabled that takes interpretation challenges like context and combination of diverse types of structured objectes into account. The framework is tested with the interpretation of terrestrial images of man-made structures.
منابع مشابه
What Can Casual Walkers Tell Us About The 3D Scene?
An approach for incremental learning of the 3D scene from a single static video camera is presented in this paper. In particular, we exploit the presence of casual people walking in the scene to infer relative depth, learn shadows, and segment the critical ground structure. Considering that this type of video data is so ubiquitous, this work provides an important step towards 3D scene analysis ...
متن کاملAutomatic Interpretation of UltraCam Imagery by Combination of Support Vector Machine and Knowledge-based Systems
With the development of digital sensors, an increasing number of high-resolution images are available. Interpretation of these images is not possible manually, which necessitates seeking for practical, fast and automatic solutions to solve the environmental and location-based management problems. The land cover classification using high-resolution imagery is a difficult process because of the c...
متن کاملAN-EUL method for automatic interpretation of potential field data in unexploded ordnances (UXO) detection
We have applied an automatic interpretation method of potential data called AN-EUL in unexploded ordnance (UXO) prospective which is indeed a combination of the analytic signal and the Euler deconvolution approaches. The method can be applied for both magnetic and gravity data as well for gradient surveys based upon the concept of the structural index (SI) of a potential anomaly which is relate...
متن کاملScene Interpretation for Lifelong Robot Learning
Integration of continual planning, monitoring, reasoning and lifelong experimental learning is necessary for a robot to deal with failures by gaining experience through incremental learning and using this experience in its future tasks. In this paper, we propose a scene interpretation system combining 3D object recognition and scene segmentation in order to maintain a consistent world model inv...
متن کاملAutomatic Learning of Conceptual Knowledge in Image Sequences for Human Behavior Interpretation
This work describes an approach for the interpretation and explanation of human behavior in image sequences, within the context of a Cognitive Vision System. The information source is the geometrical data obtained by applying tracking algorithms to an image sequence, which is used to generate conceptual data. The spatial characteristics of the scene are automatically extracted from the resuling...
متن کامل