config.yaml

# ==============================================================================
# MIR Pipeline Configuration Manifest
# ==============================================================================
# This configuration file defines the runtime parameters for the Automated
# Temporal Segmentation and Semantic Annotation pipeline. It provides the
# necessary credentials for the ACRCloud identification service and the
# constraints for the dynamic programming optimization engine.
# ==============================================================================

# ------------------------------------------------------------------------------
# ACRCloud Authentication Credentials
# ------------------------------------------------------------------------------
# The endpoint host for the ACRCloud V1 Identify API (e.g., identify-us-west-2.acrcloud.com).
# Must not include the protocol scheme (https://) or path (/v1/identify).
acr_host: "identify-eu-west-1.acrcloud.com"

# The access key provided by the ACRCloud console for the specific project.
acr_key: "your_access_key_here"

# The access secret used for HMAC-SHA1 signature generation.
acr_secret: "your_access_secret_here"

# ------------------------------------------------------------------------------
# Filesystem & Output Management
# ------------------------------------------------------------------------------
# The absolute or relative path to the directory where intermediate assets
# (downloaded audio, WAV conversions) and final outputs (snippets, JSON metadata)
# will be persisted. The pipeline will create this directory if it does not exist.
output_dir: "./mir_workspace"

# ------------------------------------------------------------------------------
# Segmentation Optimization Constraints
# ------------------------------------------------------------------------------
# The lower bound for valid segment duration in seconds.
# Used to prune the search space in the dynamic programming algorithm.
min_dur: 30.0

# The upper bound for valid segment duration in seconds.
# Used to enforce granularity and prevent under-segmentation of long mixes.
max_dur: 600.0

# ------------------------------------------------------------------------------
# System Observability
# ------------------------------------------------------------------------------
# Boolean flag to enable verbose diagnostic logging.
# If true, the pipeline emits detailed logs regarding DSP feature values,
# peak-picking thresholds, and API latency metrics.
debug_mode: true