Bio-Inspired Computing Applications 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Introduction Overview Four applications for pfrags: Audio Streaming Holistic Image Storage Surface Bus Image Segmentation Discussion/Summary 1/18/2019 Bio-Inspired Computing
Streaming Audio on a paintable Goal: Store packetized data in a particle RAM Problems: Transmission of data Storage Characteristics Retrieval 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Streaming Audio 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Streaming Audio Representation: each audio packet -> a Carrier pfrag Transport governed by migration strategy of the Carrier. Storage: the Carriers distribute uniformly in the diffusion mode. 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Streaming Stage 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Steady State 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Retrieval The output portal sends a CallBackGradient pfrag – radiates a gradient field. Contains info like ID of audio stream, “active times”, distances. Uses active time to decide what to do – 3 rules on Page 102. 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Retrieval 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Streaming audio Characterisitics: Shuttle mode playback Ubiquitous table of contents Fault tolerance No topology dependence 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Holistic Data Storage 1/18/2019 Bio-Inspired Computing
Holistic Image Storage Goal: Store a digitized image as a 2-D memory, minimizing the loss of clarity/sharpness even when a great deal of the information is not available. Duplication of the lowest frequency coefficients obtained upon transformation ensures a blurred image on reconstruction. But the size may still decrease…. 1/18/2019 Bio-Inspired Computing
Holistic Image Storage 1/18/2019 Bio-Inspired Computing
Holistic Image Storage Carriers and Transform pfrags Transform applies a “block frequency transformation” to produce a 3-level hierarchy. Output is 10 subbands – go to the carriers. Carrier splits into 9 mini-carriers. Each mini-C has 1 lowest frequency and 1 of the 9 other high frequencies. 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Image Representation 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Input –> Output 1/18/2019 Bio-Inspired Computing
Holistic Image Storage Through experiments, it is seen that: Successful decoding of images is possible. The more the number of packets the better. Multiple I/O’s can be incorporated. Discussion: Passing images to HP Additional intelligence into Carriers Hierarchical representation example. 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Surface Bus Table top containing: Devices – having short range wireless links – also called pico-nets. Particles – device transceivers contact particles in vicinity. How can they communicate with each other? comm. between external devices computation on transmitted data by particles 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Surface Bus Use channel operator and Buoy pfrag 2 regions of the ensemble Peers and portals Each peer has a unique ID. On table contact, it transmits this ID via signature Gradient. Several geometry criteria (Pg 116). 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Portal Geometry 1/18/2019 Bio-Inspired Computing
Surface Bus: Buoy pfrags 1/18/2019 Bio-Inspired Computing
Surface Bus: Buoy pfrags Need for a Buoy – to attain a finer degree of control in peer vicinity. Peers deploy a set of B pfrags upon initialization. Build the path for communication from peers. 1/18/2019 Bio-Inspired Computing
Surface Bus: Peer 2 Peer link 1/18/2019 Bio-Inspired Computing
Surface Bus: open and closed rings 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Surface Bus The purpose was to illustrate how even a simple geometry estimation can underlie a broadly useful functionality. Improvements: Conformally wrap the Co-ordinate operator More sophisticated use of fields to confine the outer ring of the table. 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Image Segmentation 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Image Segmentation 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Image Sampling 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Image Segmentation 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Image Segmentation 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Image Segmentation 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Discussion Evaluation of the programming model. Is it useful? Is it optimized? Complexity – determines scaling limits of engineered systems. Self-organization – powerful tool. 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Discussion 1/18/2019 Bio-Inspired Computing
Bio-Inspired Computing Summary We looked at 4 applications of the paintable. All these were simulated on the Psim. Questions? 1/18/2019 Bio-Inspired Computing