DIMMI - AI Companion

📋Activity Log

▾

System Ready. DIMMI initialized and awaiting training input.

⚡Actions

▾

🔊Sound Recognition

▾

Match Sensitivity: 70%

Loose Strict

Fine-Tune Recognition

Min Volume 300

How loud a sound must be. Higher = ignores quiet sounds, only reacts to loud ones.

Min Peak 50

Requires a distinct frequency peak. Higher = needs clearer tones, rejects noise.

Cooldown 3.0s

Wait time between recognitions. Higher = less responsive, fewer false triggers.

Frequency Tolerance 3 bands

How close the dominant frequency must match. Lower = stricter pitch matching.

Require Frequency Match

Only match sounds with similar pitch. Highly recommended!

Sound Signatures

No sound signatures recorded yet

🗣️Speech Recognition

▾

Add Trigger Word or Phrase

Single words AND multi-word phrases both work! Phrases are detected when spoken naturally.

Triggers & Learned Words 0

No words learned yet. Teach DIMMI to speak!

Click any word/phrase to assign a behavior = phrase = has behavior

🧠Sound Memory

▾

DIMMI remembers sounds heard in Mimic mode and plays them back randomly - like a real parrot building vocabulary!

Remember Sounds Combine Sounds

Mimic Mode Playback Chance 30%

How often DIMMI randomly says a remembered sound while in Mimic mode (0% = never, 100% = always after echo)

Memory Settings

Pitch Variation 20%

How much the pitch varies when playing back

Speed Variation 15%

How much the speed varies when playing back

Max Memories 20

Maximum sound clips to remember (oldest deleted first)

Min Length 0.5s

Minimum sound length to remember

Remembered Sounds (0)

Tap to preview, to make it a sound trigger for behaviors

No sounds remembered yet. Enable Mimic mode to start building vocabulary!

👁️Vision Detection

▾

DIMMI can detect faces, expressions, and hand gestures - and react to them!

😊 Face + Expressions 🖐️ Hand Gestures

😊 Expression → DIMMI Reactions

When DIMMI detects your facial expression, it will react! Customize below:

🖐️ Gesture → DIMMI Reactions

Show hand gestures to the camera and DIMMI will respond!

👍

Thumbs Up

✌️

Peace

🖐️

Open Palm

👆

Pointing

✊

Fist

🤘

Rock

👁️ Eye/Wink → DIMMI Reactions

Enable Eye/Wink Detection Uses face landmarks

DIMMI detects winks, blinks, and eye states using Eye Aspect Ratio (EAR) analysis!

😉

Left Wink

😜

Right Wink

😌

Long Blink

😴

Eyes Closed

🔗 Combo Reactions (Multi-Trigger)

Create powerful combos! Trigger actions when multiple things are detected together (e.g., smile + thumbs up = celebrate)

🎯 Motion Pattern Mappings

Configure DIMMI reactions to motion patterns (swipes, speed, direction)

👈

Left

👉

Right

⚡

Fast

🐢

Slow

🔄

Circle

⚙️ Advanced Vision Settings

Show Face Landmarks Detect Multiple Faces

Sensitivity 5

Speed (ms) 400

💤Idle Behavior

▾

Idle Sounds Idle Movements

Idle Settings

Sound Frequency 20s

Average time between idle sounds

Sound Types:

Chirps Whistles Purrs Squawks Clicks Trills Boops

⚙️Behavior Settings

▾

Control how DIMMI moves and reacts to you

🎯 Cursor Attraction

DIMMI moves toward cursor when happy, away when sad

Attraction Intensity 50%

Reaction Range 300px

Let mood affect direction (happy=toward, sad=away)

🔄 Idle Movement

DIMMI gently floats and wanders when not being interacted with

Movement Speed 50%

Idle Vocalization Rate 30%

? Energy Settings

Energy Decay Rate 50%

How fast energy drops when not interacting

Energy Boost Amount 10

Base energy gained per interaction

🔈 Sound Settings

Randomize sounds (different each time)

Pitch Variation 20%

🎵 Sound Library (click to preview)

Tap a sound to preview. These are the AI sounds DIMMI uses:

Sound Assignments:
• Click on stage: blip, chime
• Pet DIMMI: hum, happy
• Encourage (double-click): happy, confirm
• Idle when happy: chime, happy
• Idle when sad: sad, bored

🔗Server Connection

▾

Connect to the server to save and sync your DIMMI's learned data

Enter password to save to server