Pulse Latency Encoding for Invariant Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object recognition techniques in computer vision are inefficient due to the need for multiple filters to account for various object parameters like size, orientation, and location, leading to high complexity and cost, especially as the number of objects increases.
Innovation Solution
The development of an apparatus and method that encodes object information into a pattern of pulse latencies, where the identity of an object is encoded into the relative latencies of pulses and parameters like position, size, and orientation are encoded into a common group delay, allowing for invariant recognition across changes in these parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a bank of filters is used for matched filter approach to detect objects at multiple locations and orientations, then object detection capability is improved, but device complexity and cost increase significantly
Solution Approach 1:
The patent changes the parameter encoding approach by encoding object identity into pulse latency patterns rather than using separate filters for each object parameter combination. This allows a single filter bank to detect multiple object types and parameters by interpreting temporal patterns, dramatically reducing the number of required filters while maintaining detection capability across multiple locations, orientations, and object types
Solution Approach 2:
The patent adds a temporal dimension to object detection by using pulse latency (time) as an additional encoding dimension. Instead of requiring separate spatial filters for each object type and location, the system encodes object identity in the timing of pulses, allowing a single spatial filter to handle multiple object types through temporal discrimination, thereby reducing overall system complexity
2Measurement precision
If multiple filters are deployed for different object parameters, then recognition accuracy is maintained, but processing time and computational cost increase
Solution Approach 1:
The patent performs preliminary encoding of object identity into pulse latency patterns before detection. This pre-encoding allows the system to process multiple object types simultaneously through a single filter bank, as the temporal patterns are already prepared and differentiated, eliminating the need for sequential processing through multiple filters and thereby reducing processing time while maintaining accuracy
3Device complexity
If a single detector is used with pulse latency encoding, then device complexity is reduced, but the ability to distinguish objects with different parameters may be compromised
Solution Approach 1:
The patent introduces pulse latency patterns as an intermediary encoding mechanism between the object and the detector. Instead of requiring multiple detectors to directly distinguish different object parameters, the system uses temporal pulse patterns as a mediator that carries object identity information, allowing a single detector to differentiate multiple object types by analyzing the timing patterns of incoming pulses
Data Source
AI summary
Object recognition apparatus and methods useful for extracting information from sensory input. In one embodiment, the input signal is representative of an element of an image, and the extracted information is encoded in a pulsed output signal. The information is encoded in one variant as a pattern of pulse latencies relative to an occurrence of a temporal event; e.g., the appearance of a new visual frame or movement of the image. The pattern of pulses advantageously is substantially insensitive to such image parameters as size, position, and orientation, so the image identity can be readily decoded. The size, position, and rotation affect the timing of occurrence of the pattern relative to the event; hence, changing the image size or position will not change the pattern of relative pulse latencies but will shift it in time, e.g., will advance or delay its occurrence.