Making a Case for Learning Motion Representations with Phase
نویسندگان
چکیده
This work advocates Eulerian motion representation learning over the current standard Lagrangian optical flow model. Eulerian motion is well captured by using phase, as obtained by decomposing the image through a complexsteerable pyramid. We discuss the gain of Eulerian motion in a set of practical use cases: (i) action recognition, (ii) motion prediction in static images, (iii) motion transfer in static images and, (iv) motion transfer in video. For each task we motivate the phase-based direction and provide a possible approach.
منابع مشابه
A novel fuzzy multi-criteria decision-making methodology based upon the spherical fuzzy sets with a real case study
The choice of roll stabilization system is critical for many types of ships. For warships where operational activities are fast and the concept of time is very effective, determining the most appropriate of these systems is of particular importance. Some operations, such as the landing of the helicopter on board, are critical for naval ships. Unwanted rolling motion makes this difficult. In add...
متن کاملTBM Tunneling Construction Time with Respect to Learning Phase Period and Normal Phase Period
In every tunnel boring machine (TBM) tunneling project, there is an initial low production phase so-called the Learning Phase Period (LPP), in which low utilization is experienced and the operational parameters are adjusted to match the working conditions. LPP can be crucial in scheduling and evaluating the final project time and cost, especially for short tunnels for which it may constitute a ...
متن کاملAn Innovative Simple Test Circuit for Single-Phase Short Circuit Making Test of High-Voltage Switching Devices
Nowadays, high-voltage circuit breakers have reached such high short-circuit capabilities that testing them under the full rated voltage is generally not possible with direct tests, and they are conducted by using the synthetic test methods. Although the phenomena associated with making tests is of particular importance especially in case of load break switches, but making tests are rather d...
متن کاملEMG-based wrist gesture recognition using a convolutional neural network
Background: Deep learning has revolutionized artificial intelligence and has transformed many fields. It allows processing high-dimensional data (such as signals or images) without the need for feature engineering. The aim of this research is to develop a deep learning-based system to decode motor intent from electromyogram (EMG) signals. Methods: A myoelectric system based on convolutional ne...
متن کاملLearning Representations of Animated Motion Sequences - A Neural Model
The detection and categorization of animate motions is a crucial task underlying social interaction and perceptual decisionmaking. Neural representations of perceived animate objects are built in the primate cortical region STS which is a region of convergent input from intermediate level form and motion representations. Populations of STS cells exist which are selectively responsive to specifi...
متن کامل