Protocol

Mouse whisker object-recognition lick task

Head-fixed mice actively sample a presented object with a trimmed row of whiskers and report either object shape or object presence with left/right lick responses. The linked dataset includes both shape-discrimination and shape-detection task rules.

Task structure

Stimulus, choice, feedback, and training as declared in the protocol record.

Stimulus

evidence_schedule
A motorized object moves into whisker range and stops at a close, medium, or far position; mice actively whisk to contact the shape before choosing.
evidence_type
static
modalities
somatosensory
notes
—
units
categorical shape, position, seconds
variables
object shape, object presence, object position, whisker contact timing

Choice

action_mapping
In shape-discrimination sessions, lick side maps to concave versus convex shape; in shape-detection sessions, lick side maps to object present versus catch/no-object trials.
alternatives
left lick, right lick
choice_type
mixed
notes
—
response_modalities
lick

Feedback

feedback_type
reward
notes
—
penalty
—
reward
Water reward for correct lick choice.

Training

notes
—
stages
shaping, shape detection or discrimination, neural recording sessions

Trial timing

  1. object_approach
    approximately 2 s before response window

    Object moves toward the whiskers.

  2. sampling
    protocol-specific

    Mouse actively whisks and may contact the object.

  3. response_window
    response-limited

    Mouse reports the object class with a lick response.

Vertical slices

Expected analyses

  • accuracy by shape and object position
  • hit and false-alarm rates for detection sessions
  • touch-aligned choice and neural summaries
  • whisker-contact feature extraction

Interpretive claims

  • tactile object recognition
    Confidence: high · ref.rodgers-2022-object-recognition

    Shape identity, object presence, and active-contact kinematics are separable operational variables and should not be collapsed into a single tactile evidence scalar prematurely.

Open questions

  • Decide whether detection and discrimination sessions should split into two concrete protocols after multi-session curation.
  • Extend the optional contact-count fields into a richer per-whisker kinematics schema.