ClassifierPipeline » History » Version 2
S. Wrede, 10/02/2012 04:42 PM
1 | 1 | J. Moringen | h1. Classifier Pipeline |
---|---|---|---|
2 | 1 | J. Moringen | |
3 | 1 | J. Moringen | h2. Structuring Ideas |
4 | 1 | J. Moringen | |
5 | 1 | J. Moringen | * Data-flow layer + control layer |
6 | 1 | J. Moringen | ** E.g. central controller manages component states according to some protocol/rules |
7 | 1 | J. Moringen | |
8 | 1 | J. Moringen | * Alternative: Subgraph lifecycles |
9 | 1 | J. Moringen | ** E.g. learning subgraph between data buffer and learner) |
10 | 1 | J. Moringen | |
11 | 2 | S. Wrede | h2. Requirements |
12 | 1 | J. Moringen | |
13 | 2 | S. Wrede | h3. Training |
14 | 2 | S. Wrede | |
15 | 1 | J. Moringen | * Sequential processing: data source, keypoint, descriptor, feature |
16 | 1 | J. Moringen | * Processing Strategies (in CCA speech) |
17 | 1 | J. Moringen | ** First possibility |
18 | 1 | J. Moringen | *** As fast as possible |
19 | 1 | J. Moringen | *** Flow control is necessary |
20 | 1 | J. Moringen | *** Usually, all inputs are processed |
21 | 1 | J. Moringen | ** Second possibility |
22 | 1 | J. Moringen | *** "Live" data source (i.e. camera) |
23 | 1 | J. Moringen | * Multiple phases |
24 | 1 | J. Moringen | ** Collection |
25 | 1 | J. Moringen | ** Actual training |
26 | 1 | J. Moringen | ** Model is transmitted from learner to classifier |
27 | 1 | J. Moringen | |
28 | 2 | S. Wrede | h3. Classification |
29 | 1 | J. Moringen | |
30 | 1 | J. Moringen | * Static, linear data flow: source, keypoint, descriptor, feature, classifier |
31 | 1 | J. Moringen | * No buffering/persistence required |
32 | 2 | S. Wrede | |
33 | 2 | S. Wrede | h2. Non-functional Requirements |
34 | 2 | S. Wrede | |
35 | 2 | S. Wrede | * Parallel / distributed processing |
36 | 2 | S. Wrede | * Visualization (when and what) |
37 | 2 | S. Wrede | * Model description / specification (do skeletons exist?) |
38 | 2 | S. Wrede | |
39 | 2 | S. Wrede | h2. Related work |
40 | 2 | S. Wrede | |
41 | 2 | S. Wrede | * gstreamer |
42 | 2 | S. Wrede | ** usability is an issue |
43 | 2 | S. Wrede | * IceWing |
44 | 2 | S. Wrede | ** data flow specification |
45 | 2 | S. Wrede | ** parallel processing (?) |
46 | 2 | S. Wrede | ** distributed processing (?) |
47 | 2 | S. Wrede | ** support for non-image data is not existent |
48 | 2 | S. Wrede | ** usability (?) |
49 | 2 | S. Wrede | ** no support for collection, training, classification phases / states |