AI-SPEAK / LipReadingApp / config.py
config.py
Raw
# UI / camera / extraction parameters
IMG_WIDTH = 100
IMG_HEIGHT = 50
MAX_NUM_FRAMES = 150

RECORD_SECONDS = 6.0
MIN_DET_CONF = 0.5
MIN_TRACK_CONF = 0.5
LIP_PADDING = 10
CAM_INDEX = 0

# Landmarks for outer lips (MediaPipe indices)
OUTER_LIPS = [
    61, 185, 40, 39, 37, 0,   267, 269, 270, 409,
    291, 375, 321, 405, 314, 17,  84,  181, 91,
    146, 61
]

MODEL_PATH = r"./model/"

NORM_STATS_1080x1920_JSON = r"./data/global_stats_base.json"  # e.g. r"d:/.../global_stats_1080x1920.json"
NORM_STATS_OTHER_JSON     = r"./data/global_stats.json"  # e.g. r"d:/.../global_stats_other.json"

LIP_PAD_W_RATIO = 0.05
LIP_PAD_H_RATIO = 0.1

PRESCALE_RANGE_1080x1920 = (0.60, 0.70)

# For all other inputs:
PRESCALE_RANGE_OTHER = (1.0, 1.0)

PRESCALE_SEED = None