All system configurations can be shown and edited from the "Arena Configurations" panel.
To open this panel, you can click the cog icon (⚙) on the top-left side of PreyTouch UI.
There are 4 JSON configuration files in the system, that can be edited in PreyTocuh UI:
- cameras - Specify all the cameras in the system. From this window you can also scan the cameras in the system (only FLIR and Allied-Vision), and also add a new camera using a simple form.
- periphery - Specify all the peripheral devices
- predict
- agent
- camera name
- id - serial number of the camera
- module - currently supported modules: flir, allied_vision
- exposure - exposure in microseconds
- image_size - image size in pixels
- output_dir - leave blank. This is used by the video writing module.
- always_on - if true, the scheduler tries to start the camera if off.
- fps - If set, the camera takes frames according to this value.
- trigger_source - if set, camera takes frames according to trigger
- writing_fps - FPS for video writing, can be lower than camera's fps
- is_color - set true, if the camera's frames are with 3 channels
- predictors - specify real-time predictors
- predictor name (must be configured in predict_config)
- image_size - image size provided to the predictor
- mode - can be "experiment" (run only during experiments), "no_experiment" (run only when there's no experiment running) or "always" (run always).
- movement_type - specify the specific movement types in which the predictor is allowed to run.
- predictor name (must be configured in predict_config)
{
"front": {
"id": 19506475,
"module": "flir",
"exposure": 7000,
"image_size": [
1080,
1440
],
"output_dir": null,
"always_on": true,
"fps": 60,
"writing_fps": 60,
"is_color": false,
"predictors": {
"tongue_out": {
"image_size": [
1080,
1440
],
"mode": "experiment",
"movement_type": [
"jump_up",
"accelerate",
"circle_accelerate"
]
}
}
},
"top": {
"id": 19506455,
"module": "flir",
"exposure": 4000,
"image_size": [
1080,
1440
],
"output_dir": null,
"always_on": true,
"fps": 10,
"writing_fps": 10,
"is_color": false,
"mode": "tracking",
"predictors": {}
}
}- arduino name
- allow_get - (bool) allow get status
- fqbn - Fully Qualified Board Name
- interfaces - list of devices
- name - device name
- pin - single arduino pin
- pins - list of pins
- type - can be line, feeder, dallas_temperature, trigger
- serial_number - Arduino serial number
{
"arena": {
"allow_get": true,
"fqbn": "arduino:megaavr:nona4809",
"interfaces": [
{
"command": "dispense",
"name": "Feeder 1",
"pins": [
10,
9,
8,
7
],
"type": "feeder",
"order": 1
},
{
"command": "dispense",
"name": "Feeder 2",
"pins": [
14,
15,
16,
17
],
"type": "feeder",
"order": 2
},
{
"name": "day_lights",
"pin": 2,
"type": "line"
},
{
"name": "IR_lights",
"pin": 13,
"type": "line"
},
{
"name": "Temp",
"pin": 5,
"type": "dallas_temperature"
}
],
"serial_number": "0E4273DD51534C5036202020FF072035"
},
"camera trigger": {
"allow_get": true,
"fqbn": "arduino:megaavr:nona4809",
"interfaces": [
{
"name": "Camera Trigger",
"pin": 12,
"pulse_len": 33,
"pulse_width": 0.7,
"serial_trigger": false,
"type": "trigger"
}
],
"serial_number": "91CEA40751534C5036202020FF07444F"
}
}
- predictor name
- predictor name - predictor class name
- model_path - path to model
- threshold - predictions thresholds
- all the rest are kwargs for the predictor class
{
"deeplabcut": {
"predictor_name": "DLCPose",
"model_path": "/data/Pogona_Pursuit/output/models/deeplabcut/front_head_only_resnet_152",
"bodyparts": [
"nose",
"right_ear",
"left_ear"
],
"threshold": 0.5
},
"pogona_head": {
"model_path": "",
"predictor_name": "PogonaHead",
"threshold": 0.5
},
"tongue_out": {
"model_path": "/data/Pogona_Pursuit/output/models/tongue_out/20230518_145847",
"dataset_path": "/data/Pogona_Pursuit/output/datasets/pogona_tongue/dataset",
"save_predicted_path": "/data/Pogona_Pursuit/output/datasets/pogona_tongue/predicted/tongues",
"predictor_name": "TongueOutAnalyzer",
"threshold": 0.7,
"image_size": [
550,
1000
],
"prediction_stack_duration": 0.25,
"tongue_action_timeout": 0.5,
"num_tongues_in_stack": 6
}
}- default_struct (default experiment parameters)
- time_between_blocks - time in seconds between consecutive blocks.
- extra_time_recording - time in seconds before and after the experiment for extra recording. The cameras record, but no trials are starting.
- num_blocks - Number of blocks in each experiment.
- is_identical_blocks - in case there are multiple blocks, make them all identical.
- is_test - test experiment, app is started on the configured TEST_SCREEN and no rewards are given.
- reward_bugs - specify the bugs which trigger reward. If null, all bugs are being rewarded.
- background_color - app background color in hex.
- exit_hole: can be "left", "right" or "random"
- reward_any_touch_prob - probability to get reward even if missed.
- cameras - specify all the cameras as keys and put any needed camera argument below (see example)
- blocks - default block parameters:
- num_trials - number of trials in each blocks
- trial_duration - default trial duration in seconds
- iti - inter trial interval in seconds
- block_type - bugs or media
- notes - notes to be saved for the block
- bug_speed - default bug speed
- is_default_bug_size - use the default bug size, that specified in app the config
- bug_size - bug size in pixels
- times
- start_time - Time for the agent to start setting experiments (format: "HH:MM").
- end_time - End time for the agent (format: "HH:MM").
- time_between_experiments - Time in minutes between scheduled experiments.
- trials
{
"default_struct": {
"time_between_blocks": 180,
"extra_time_recording": 30,
"num_blocks": 1,
"is_identical_blocks": false,
"is_test": false,
"reward_bugs": null,
"background_color": "#e8eaf6",
"exit_hole": "random",
"reward_any_touch_prob": 0.1,
"cameras": {
"back": {
"is_use_predictions": true
},
"front": {
"is_use_predictions": true
}
},
"blocks": [
{
"num_trials": 10,
"trial_duration": 30,
"iti": 20,
"block_type": "bugs",
"notes": "created by agent",
"bug_speed": 5,
"movement_type": null,
"is_default_bug_size": true
}
]
},
"times": {
"start_time": "09:00",
"end_time": "18:15",
"time_between_blocks": 60
},
"trials": {
"random_low_horizontal": {
"count": {
"key": "strikes",
"amount": 40,
"per": {
"bug_speed": [
2,
4,
6,
8
]
}
},
"exit_hole": "random",
"bug_speed": "per_random",
"movement_type": "random_low_horizontal"
},
"circle": {
"count": {
"key": "strikes",
"amount": 40,
"per": {
"bug_speed": [
2,
4,
6,
8
]
}
},
"exit_hole": "random",
"bug_speed": "per_random",
"reward_any_touch_prob": 0.1,
"movement_type": "circle"
},
"circle_accelerate": {
"count": {
"key": "strikes",
"amount": 100
},
"exit_hole": "random",
"reward_any_touch_prob": 0.1,
"movement_type": "circle_accelerate",
"bug_speed": 4
},
"low_horizontal": {
"count": {
"key": "trials",
"amount": 200,
"per": {
"exit_hole": [
"bottomLeft",
"bottomRight"
]
}
},
"exit_hole": "per_ordered",
"bug_speed": 6,
"movement_type": "low_horizontal",
"reward_any_touch_prob": 0,
"num_trials": 5
}
}
}