Mental Commands Introduction
The Mental Commands detection suite evaluates a user’s real time brainwave activity to discern the user’s conscious intent to perform distinct physical actions on a real or virtual object. The detection is designed to work with up to 13 different actions: 6 directional movements (push, pull, left, right, up and down) and 6 rotations (clockwise, counter-clockwise, left, right, forward and backward) plus one additional action that exists only in the realm of the user’s imagination: disappear.
Mental Commands allows the user to choose up to 4 actions that can be recognized at any given time. The detection reports a single action or neutral (i.e. no action) at a time, along with an action power which represents the detection’s certainty that the user has entered the Mental Commands state associated with that action.
Increasing the number of concurrent actions increases the difficulty in maintaining conscious control over the Mental Command detection results. Almost all new users readily gain control over a single action quite quickly. Learning to control multiple actions typically requires practice and becomes progressively harder as additional actions are added. Although Emotiv Xavier allows a user to select up to 4 actions at a time, it is important that each user masters the use of the Mental Commands detection one action at a time, only increasing the number of concurrent actions after he has first gained confidence and accuracy with a lower number of actions.
Understanding the Mental Commands Tab Display
The Mental Commands Tab uses a virtual 3D cube to display an animated representation of the Mental detection output. This 3D cube is also used to assist the user in visualizing the intended action during the training process. The Power gauge on the top of the 3D display is an indicator of the “action power”, or relative certainty that the user is consciously visualizing the current action.
The default tab on the Mental Commands Tab is the Action tab. This tab displays information about the current state of the Mental detection and allows the user to define the current set of actions.
In order to enable the Mental Commands detection, each chosen action, plus the Neutral action, must first be trained. Alongside each currently selected action is another gauge displaying a Skill Rating. This skill rating is calculated during the training process and provides a measure of how consistently the user can mentally perform the intended action. It is necessary to train the same action at least two times before the action skill is updated. The Overall Skill Rating is simply the average of all the individual action skills and can be used as a general measure of the user’s skill with the selected set of actions and existing training data.
A green checkmark is used to indicate that the corresponding action has been trained; a red X indicates a lack of training data. Remember, in order for the Mental Commands detection to be activated, all actions, plus Neutral (the user’s background Mental Commands state) must be trained.
Use the Add, Remove, and Edit push buttons to modify the number and type of enabled actions.
Mental Commands Training
The Mental Commands training process enables the EPOC to analyze your brainwaves and develop a personalized signature which corresponds to each particular action, as well as the background state, or “neutral”. As the Insight learns and refines the signatures for each of the actions, as well as neutral, detections become more precise and easier to perform.
The Training tab contains the user interface controls that support the Mental Commands training process. The training process consists of three steps:
First, select an action from the dropdown list. Actions that have already been trained are paired with a green checkmark; actions with no training data are paired with a red X. Only actions that have been selected on the Action tab are available for training. The default action is one without training data; otherwise the default action is “Neutral.”
Next, when you are ready to begin imagining or visualizing the action you wish to train, press the Start Training button. During the training process it is very important to maintain your cognitive focus for the duration of the training period (currently 8 seconds). Physical gestures, such as pushing an imaginary object with one hand, may be used to heighten your focus on the intended action, but are not required. You should also refrain from making substantial head movements or dramatic Facial Expressions during the training period, as these actions can interfere with the recorded EEG signal.
Initially, the cube on screen will not move, as the system has not yet acquired the training data necessary to construct a personalized signature for the current set of actions. After Neutral and each enabled action have been trained at least once, the Mental Commands detection is activated and the cube will respond to the Mental Commands detection, and your Mental Commands control, in real time.
Some users will find it easier to maintain the necessary Mental Commands focus if the cube is automatically animated to perform the intended action as a visualization aid during training. If you think you will benefit from this, then you may select the Animate model according to training action checkbox. Otherwise, the cube will remain stationary or, if you have already supplied training data and the detection is active, will be animated by the current detection results for the action being trained, while you supply new training data.
Finally, you are prompted to accept or reject the training recording. Ideal Mental Commands detection performance is typically achieved by supplying consistent training data (i.e. a consistent Mental Commands visualization on the part of the user) across several training sessions for each enabled action. The ability to reject the last training recording allows you to decide whether you were able to remain mentally focused on the appropriate action during the last training session. Alternatively, you may press the Abort Training button to abort the training recording if you are interrupted, become distracted, or notice problems with the neuroheadset contact quality indicators during the recording.
A training session is automatically discarded if the wireless signal strength or EEG signal quality is poor for a significant portion of the training period. A notification will be displayed to the user if this has occurred.
The Neutral “action” refers to the user’s passive Mental Commands state; one that isn’t associated with any of the selected Mental Commands actions. While training Neutral, you should enter a Mental Commands state that doesn’t involve the other Mental Commands actions. Typically this means engaging in passive Mental Commands activities such as reading or just relaxing. However, to minimize “false-positive” Mental Commands action results (i.e. incorrect reporting of unintended actions), it may also be helpful to emulate other Mental Commands states and Facial Expressions that are likely to be encountered in the application context and environment in which you’ll be using Mental. For many users, providing more Neutral training data will result in better overall Mental performance. In order to facilitate the acquisition of this data, and because most users have an easier time maintaining a relaxed mental state for longer periods of time, the Training tab provides a second method for training Neutral.
The Record Neutral button allows you to record up to 30 seconds of Neutral training data. The recording will automatically finish, and the Mental signature will be rebuilt, after 30 seconds, or you may press the Stop Recording button at any time you feel sufficient data has been collected. Note: at least 6 seconds of recorded data will be required to update the signature with data collected using the Record Neutral button.
Clear Training Data Button
Occasionally, you may find that a particular trained action doesn’t work as well as it once did. This may indicate that the training data used to construct your personalized Mental signature was “contaminated” by a more recent, inconsistent training session or that some characteristics of your brainwaves have changed over time. It may also happen that you wish to change the mental imagery or technique that you associate with a particular action. In either situation, you can use the Clear Training Data button to delete the training data for the selected action. Keep in mind that doing so will disable the Mental Commands detection until new training data has been recorded for this action.
Advanced Mental Commands Options
The Advanced tab exposes settings and controls that allow you to customize the behavior of the Mental Commands detection. By default, the Mental Commands detection is pre-configured in a manner that produces the best results for the largest population of users and environments. It is strongly recommended that you only change these settings with the guidance of Emotiv personnel who are better acquainted with the internal details of the Emotiv EmoEngine.
Mental Commands Tips
Cognitive dexterity with Mental Commands is a skill that will improve over time. As you learn to train distinct, reproducible Mental Commands states for each action, the detection becomes increasingly precise. Most users typically achieve their best results after training each action several times. Overtraining can sometimes produce a decrease in accuracy – although this may also indicate a lack of consistency and cognitive fatigue. Practice and experience will help determine the ideal amount of training required for each individual user.
If it becomes hard for you to return to neutral (i.e. to stop the cube from moving) you should try refreshing your Mental Commands state by momentarily shifting your focus away from the screen and relaxing. It is easy to become immersed in the experience and to have the Mental Commands actions at the “front of your mind” while trying to be neutral.
Successful training relies on consistency and focus. For best results, you must perform the intended action continuously for the full training period. It is common for novice users to become distracted at some point during the training period and then cognitively restart an action, but this practice will result in poorer results than training with a cognitive focus that spans the entire training period.
A short latency, of up to two seconds, in the initiation and cessation of the cube’s animated action on screen is typical.
Mental Commands Settings
Settings Tab is a new feature of Emotiv Xavier. You can use your EPOC headset to customize and control imported 3D Objects.
First you have to Import your model, edit the scale to suit, and arrange the background to create custom context. Then you can use the EPOC Mental Commands detection suite to control any 3D Object in the Xavier SDK window.
The MODEL CONTROL tab allows you to set the Background, Set Frame for the usual Cube. You can click Set Default button to reset default state. The window on the right allow you preview settings relate to display, then you click Apply button to display on Cube on the left window.
There are two types display for cube: 2D and 3D cube. You just need to click back and forth between 2D and 3D will look clearer of the differences between 2D and 3D picture.
Emotiv Xavier preset for you some model 3D to use. There are two 3D Models: Car and Box in this program for you may be selected.
Select the 3D Object from the Option tab and use the EPOC headset to control them in the Mental Commands Detection Suite.
You can combine multiple actions to control the model (Rotate Clockwise and Left/Right) or (Rotate Reverse vs Left/Right).
IMPORT PAGE allows the user to set up the context with your own 3D models which can be manipulated using your Mental Commands. This tab allows you to experience some of the other models with the default interface of the program. Here you can test the ability to sort and decorate the model.
Step 1: Import model
Click Import Model button, then you navigate to the folder where you stored 3DS Model (.3ds) file or Emotiv Model Setting (.EMS) file if it is available.
Step 2: Place the model in your desired position then add Texture, edit Scale, Rotate and Position it.
Step 3: Repeat step 1 and step 2 until your context is completed. If you are not satisfied with any models, you can remove it by choosing Delete button.
Step 4: Select any model in that context. You can use the EPOC headset to control model. You need to click the Command Object button to start. The chosen Model is transferred the respective action in Action Tab. You could click the window to the left of the desktop, then drag the mouse to see the change of rotation angle of the Model, which helps you have a new view of your chosen model.
Step 5: Import Model and Export Setting
Export Setting: After you have created a context you like and want to save it to be able to use it again next time you can use the Export Settings. This file is saved in the format of a .ems file. With the certain context you have created, Emotiv Xavier allows you to keep still the Position, Texture, and Scale of the model that can then be reloaded at a later time.
IMPORT MODEL: If you want to reload a previously used context, you can reload it from the .EMS file. Then the whole context of the file will be will be reloaded to the program.
The file idea1.ems is provided as an example.
This program also supports the following tools for manipulating the design:
Key G: enable/disable Grid
Key V: show camera Viewport
Key S: Pause/Run Action
Mental Commands Challenge
The Challenge tab allows you to challenge with cube model and submit your score to Emotiv’s website. You need to select at least one trained action to begin challenge.
You can only use the challenge function when at least half the sensors are "green" showing the best connection to ensure accurate data collection. If there is no trained action, it will be necessary to first train at least one action to continue. When the power bar at the top of the window changes, data will start to collect, and a score will be generated. Score displays your current score, and Time Remaining presents the remaining time in your challenge session. The maximum time duration for each session is 60 seconds. You can also initiate a challenge for a shorter time period by clicking the "END NOW" button.
Click the START button to begin playing. You can click Esc keyboard to disable fullscreen mode, and the END NOW button to finish challenge. After you finish playing, click the SUBMIT SCORE button to submit your score to Emotiv’s website under your username. The HIGHSCORE button displays the highest score online. To view your Score online, you can click View Online Leaderboard in HighScore or View online user percentile rank above Score table.
The right-hand table displays Scores recorded only on your local machine.