Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/sudochris/evolution-evaluation
https://github.com/sudochris/evolution-evaluation
Last synced: about 1 month ago
JSON representation
- Host: GitHub
- URL: https://github.com/sudochris/evolution-evaluation
- Owner: sudochris
- License: gpl-3.0
- Created: 2021-05-30T16:03:22.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2022-03-05T21:05:26.000Z (almost 3 years ago)
- Last Synced: 2023-10-20T07:48:51.361Z (about 1 year ago)
- Language: Python
- Size: 11.8 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Evolution Evaluation
# Command line parameters
| Parameter | Description | Default |
| :----------- | :----------------------------------- | :-------------------------------------- |
| `-pf` | Path to parameters for the algorithm | data/squash/parameters.json |
| `-gf` | Path to geometry file for evolution | data/squash/geometries/squash_court.obj |
| `-erf` | Output path for evolution results | evolution_results.csv |
| `-nrf` | Output path for nlopt results | nlopt_results.csv |
| `-evolution` | Run evolution experiments | (flag) |
| `-nlopt` | Run NLOPT experiments | (flag) |
| `-headless` | Run experiments headless | (flag) |
e.g.: `python evaluation.py -erf evo_results.csv -nrf nlopt_results.csv -evolution -evolution -nlopt`
Results can be found in `data/`
> **Please Note**: A PARALELLIZED VERSION FOR PERFORMANCE BOOST IS CURRENTLY UNDER DEVELOPMENT
# Evolution Evaluation
- The `target` vector represents always the mean camera.
- The `start` vector represents the "wiggled" mean camera:
The number of wiggled parameters is set to 7, which is 46% of all 15 parameters. Every parameter has a chance to get selected for variation. The amount of wiggling is specified as follows:
| Parameter | Description | Near | Medium | Far |
| :-------- | :---------------------------: | :----: | :----: | :--: |
| fu | focal u | ±10 | ±50 | ±100 |
| fv | focal v | ±10 | ±50 | ±100 |
| cx | principal point offset x [px] | ±10 | ±50 | ±100 |
| cy | principal point offset y [px] | ±10 | ±50 | ±100 |
| tx | translation in x [m] | ±0.1 | ±0.5 | ±1.0 |
| ty | translation in y [m] | ±0.1 | ±0.5 | ±2.0 |
| tz | translation in z [m] | ±0.1 | ±0.5 | ±1.0 |
| rx | rotation in x [rad] | ±0.01 | ±0.1 | ±0.2 |
| ry | rotation in y [rad] | ±0.01 | ±0.1 | ±0.2 |
| rz | rotation in z [rad] | ±0.01 | ±0.1 | ±0.2 |
| d0 | distortion parameter | ±0.1 | ±0.25 | ±0.5 |
| d1 | distortion parameter | ±0.1 | ±0.25 | ±0.5 |
| d2 | distortion parameter | ±0.025 | ±0.1 | ±0.2 |
| d3 | distortion parameter | ±0.01 | ±0.05 | ±0.1 |
| d4 | distortion parameter | ±0.5 | ±1.5 | ±3.0 |
### Start camera specification
Given a randomly chosen wiggle vector for a specified distance type, the start is defined by:
['s_fu', 's_fv', 's_cx', 's_cy', 's_tx', 's_ty', 's_tz', 's_rx', 's_ry', 's_rz', 's_d0', 's_d1', 's_d2', 's_d3', 's_d4']
### Target camera specification
The (in a real world application unknown) target values are:
['t_fu', 't_fv', 't_cx', 't_cy', 't_tx', 't_ty', 't_tz', 't_rx', 't_ry', 't_rz' 't_d0', 't_d1', 't_d2', 't_d3', 't_d4']
> Pleas Note that this is always the mean squash camera in **our synthetic case**!
### Evolution Algorithm related variables
| Variable | Description | Variants |
| :------------- | :------------------------------------------- | :------------------------------- |
| population_fn | Strategy for creating the initial population | BoundedUniformPopulation |
| | | ValueUniformPopulation |
| fitness_fn | | DistanceMap |
| | | DistanceMapWithPunishment |
| selection_fn | | RouletteWheel |
| | | Tournament |
| | | Random |
| crossover_fn | | SinglePoint |
| | | TwoPoint |
| mutation_fn | | BoundedUniformMutation |
| | | BoundedDistributionBasedMutation |
| termination_fn | | MaxIteration |
| | | NoImprovement |
| | | FitnessReached |
| | | And |
| | | Or |
### Noise related variables
| Variable | Description | Variants |
| :--------- | :---------- | :------------ |
| noise_type | | no_noise |
| | | salt |
| | | hlines |
| | | straight_grid |
| | | angled_grid |
### Output Format
The result file is a comma separated value file with the following specification:
**Strategy bundle columns**
| Column | Description | Note |
| --------------- | ------------------------------------------- | ------------------ |
| nlopt_optimizer | nlopt optimizer used | **NLOPT ONLY** |
| population_fn | Population strategy | **EVOLUTION ONLY** |
| fitness_fn | Fitness stategy | |
| selection_fn | Selection strategy | **EVOLUTION ONLY** |
| crossover_fn | Crossover strategy | **EVOLUTION ONLY** |
| mutation_fn | Mutation strategy | **EVOLUTION ONLY** |
| termination_fn | Termination strategy | **EVOLUTION ONLY** |
| distance_type | Distance selected for the wiggle operator | |
| noise_type | Noise type | |
| noise_value | If applicable: Parameter for the noise type | |
**Camera columns**
| Column | Description |
| ------ | :----------------------------------------------------- |
| s_fu | **Start camera** fu parameter (Focal u) |
| s_fv | **Start camera** fv parameter (Focal v) |
| s_cx | **Start camera** cx parameter (Central point offset x) |
| s_cy | **Start camera** cy parameter (Central point offset y) |
| s_tx | **Start camera** tu parameter (Translation X) |
| s_ty | **Start camera** ty parameter (Translation Y) |
| s_tz | **Start camera** tz parameter (Translation Z) |
| s_rx | **Start camera** ru parameter (Rotation X) |
| s_ry | **Start camera** ry parameter (Rotation Y) |
| s_rz | **Start camera** rz parameter (Rotation Z) |
| s_d0 | **Start camera** d0 parameter (Distortion k1) |
| s_d1 | **Start camera** d1 parameter (Distortion k2) |
| s_d2 | **Start camera** d2 parameter (Distortion p1) |
| s_d3 | **Start camera** d3 parameter (Distortion p2) |
| s_d4 | **Start camera** d4 parameter (Distortion k3) |
| Column | Description |
| ------ | ------------------------------------------------------- |
| t_fu | **Target camera** fu parameter (Focal u) |
| t_fv | **Target camera** fv parameter (Focal v) |
| t_cx | **Target camera** cx parameter (Central point offset x) |
| t_cy | **Target camera** cy parameter (Central point offset y) |
| t_tx | **Target camera** tu parameter (Translation X) |
| t_ty | **Target camera** ty parameter (Translation Y) |
| t_tz | **Target camera** tz parameter (Translation Z) |
| t_rx | **Target camera** ru parameter (Rotation X) |
| t_ry | **Target camera** ry parameter (Rotation Y) |
| t_rz | **Target camera** rz parameter (Rotation Z) |
| t_d0 | **Target camera** d0 parameter (Distortion k1) |
| t_d1 | **Target camera** d1 parameter (Distortion k2) |
| t_d2 | **Target camera** d2 parameter (Distortion p1) |
| t_d3 | **Target camera** d3 parameter (Distortion p2) |
| t_d4 | **Target camera** d4 parameter (Distortion k3) |
| Column | Description |
| ------ | ------------------------------------------------------ |
| r_fu | **Resultcamera** fu parameter (Focal u) |
| r_fv | **Resultcamera** fv parameter (Focal v) |
| r_cx | **Resultcamera** cx parameter (Central point offset x) |
| r_cy | **Resultcamera** cy parameter (Central point offset y) |
| r_tx | **Resultcamera** tu parameter (Translation X) |
| r_ty | **Resultcamera** ty parameter (Translation Y) |
| r_tz | **Resultcamera** tz parameter (Translation Z) |
| r_rx | **Resultcamera** ru parameter (Rotation X) |
| r_ry | **Resultcamera** ry parameter (Rotation Y) |
| r_rz | **Resultcamera** rz parameter (Rotation Z) |
| r_d0 | **Resultcamera** d0 parameter (Distortion k1) |
| r_d1 | **Resultcamera** d1 parameter (Distortion k2) |
| r_d2 | **Resultcamera** d2 parameter (Distortion p1) |
| r_d3 | **Resultcamera** d3 parameter (Distortion p2) |
| r_d4 | **Resultcamera** d4 parameter (Distortion k3) |
**Result columns**
| Column | Description |
| ----------------------------- | ------------------------------------------------------------------------------ |
| best_fitness | Best achieved fitness |
| generations | Number of generations until convergence |
| mean_fitting_reproj_error_col | Mean reprojection error for fitting geometry (horizontal in image space) |
| mean_fitting_reproj_error_row | Mean reprojection error for fitting geometry (vertical in image space) |
| mean_dense_reproj_error_col | Mean reprojection error for dense sampled geometry (horizontal in image space) |
| mean_dense_reproj_error_row | Mean reprojection error for dense sampled geoemtry (vertical in image space) |
| mean_y0_reproj_error_col | Mean reprojection error for court floor geometry (horizontal in image space) |
| mean_y0_reproj_error_row | Mean reproejction error for court floor geometry (vertical in image space) |
### Citation
Please cite in your publications if it helps your research:
TBA
### License
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see .