Event Vision Frame Prediction for Synchronous Data Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image sensors employing event vision pixels face challenges in transmitting asynchronous event data using synchronous communications interfaces due to latency issues and frame transmission aborts caused by varying event rates, leading to data loss.
Innovation Solution
Implementing a synchronous communications transmitter that anticipates event data rates by predicting or estimating the amount of data in future frames, allowing the receiver to adjust its circuitry accordingly, and adjusting frame sizes to match anticipated data using predictor blocks and memory buffers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If asynchronous event data is transmitted through synchronous communications interfaces, then data communication is enabled, but latency variability and data loss occur due to varying event rates
Solution Approach 1:
The system performs preliminary actions by predicting future event rates and pre-adjusting frame sizes and buffer allocations before data transmission occurs. The predictor block anticipates event data rates and communicates these predictions to the transmitter in advance, allowing the system to proactively adapt to varying event rates rather than reactively responding to them, thereby preventing latency variability and data loss
Solution Approach 2:
The system implements feedback mechanisms where the predictor block continuously monitors actual event rates and compares them with predicted rates. This feedback loop allows the system to dynamically adjust frame sizes and buffer allocations based on real-time performance, improving data transmission reliability while maintaining compatibility with synchronous interfaces
2Stability of the object's composition
If frame sizes are fixed for synchronous transmission, then interface compatibility is maintained, but data loss occurs when event rates vary
Solution Approach 1:
The system applies dynamics by making frame sizes adjustable rather than fixed. The predictor block determines optimal frame sizes based on predicted event rates, and the transmitter dynamically adjusts frame sizes within allowable ranges. This dynamic adaptation allows the system to maintain synchronous interface compatibility while preventing data loss during event rate variations
Solution Approach 2:
The system changes parameters by adjusting frame size and buffer allocation based on predicted event rates. Instead of maintaining a fixed frame size, the system varies these parameters dynamically to match actual data generation rates, thereby preventing overflow and underflow conditions that cause data loss while maintaining overall transmission stability
3Loss of time
If buffer sizes are reduced to decrease latency, then transmission speed improves, but data loss increases due to event rate variations
Solution Approach 1:
The system performs preliminary action by predicting event rates and pre-configuring buffer sizes before data transmission. This allows the system to maintain smaller buffers that reduce latency while ensuring they are sized appropriately for the predicted event rate, preventing overflow and data loss without requiring large buffers that would increase latency
Data Source
AI summary
Methods for transmitting asynchronous event data via synchronous communications interfaces (and associated imaging systems) are disclosed herein. In one embodiment, an imager comprises an array of event vision pixels, and a synchronous communications transmitter configured to transmit frames of data to a synchronous communications receiver. The pixels generate event data based on activity within an external scene. The imager communicates, at a first time and to the receiver, an anticipated amount of data that will be included in a frame transmitted to the receiver at a second time. The anticipated amount of data can be based on a prediction of an amount of event data that will be generated at a future point in time for transmission to the receiver in the frame. The imager can then transmit the frame to the receiver at the second time with an amount of data corresponding to the anticipated amount of data.