Tasks #2019
Add VFOA Results to PersonHypothesis
Status: | Resolved | Start date: | 09/23/2014 | |
---|---|---|---|---|
Priority: | Normal | Due date: | ||
Assignee: | J. Wienke | % Done: | 100% | |
Category: | Type Proposal | |||
Target version: | Robotics Service Bus - rsb-0.11 |
Description
The results of the visual focus of attention are needed in the PersonHypothesis. While a HeadObject carries only the best target, the PersonHypothesis could gather all targets including their probabilities.
Associated revisions
Add attention targets to PersonHypothesis
A new repeated filed allows to express the probability that a person
focuses on individual named attention targets.
fixes #2019
Signed-off-by: Johannes Wienke <jwienke@techfak.uni-bielefeld.de>
History
#1 Updated by J. Wienke over 9 years ago
- Status changed from New to In Progress
- Assignee set to J. Wienke
- Target version set to rsb-0.11
#2 Updated by J. Wienke over 9 years ago
- File 0001-Add-attention-targets-to-PersonHypothesis.patch added
- % Done changed from 0 to 30
- better express the meaning of the new fields
- make them independent of visual focus of attention for more generality
- fix coding conventions (tabs -> spaces, newlines, constraint added, improved naming)
In this form, the additions for attention targets largely resemble the ClassificationResult type and it could be argued that the current VFOA algorithm is more or less a classification algorithm. Personally, I would, however, use the more specifically named type here to be more expressive and allow further domain-specific changes which wouldn't be possible with the general classification result type. Any opinions on that?
#3 Updated by J. Wienke over 9 years ago
Ping?
#4 Updated by T. Dankert over 9 years ago
Uh, sorry. Your suggestion would also work for me.
#5 Updated by T. Dankert over 9 years ago
- Status changed from In Progress to Resolved
- % Done changed from 30 to 100
Applied in changeset rst-proto|a45a11d397aacc3fe0f84308464aaf1e51eef665.