Python: feat(camb): Add CAMB AI connector for text-to-speech, speech-to-text, and audio services

[neilruaro-camb]

Motivation and Context

CAMB.ai provides multilingual AI-powered audio services. This PR adds a connector that integrates CAMB.ai capabilities into Semantic Kernel, enabling developers to use text-to-speech, speech-to-text, translation, voice cloning, and other audio features through the standard SK service interfaces.

Description

Adds a new camb connector under semantic_kernel/connectors/ai/camb/ with the following components:

Services:

• CambTextToAudio — TTS service supporting multiple output formats (WAV, MP3, FLAC, OGG, PCM, Mu-law) via MARS models
• CambAudioToText — Speech-to-text transcription service with async task polling

Plugin:

• CambPlugin — Kernel plugin exposing 6 additional features as kernel functions:
  • translate — Translate text between 140+ languages
  • translated_tts — Translate and synthesize speech in one call
  • clone_voice — Create custom voice clones from audio samples
  • list_voices — List available voices with metadata
  • text_to_sound — Generate sound effects/music from text descriptions
  • separate_audio — Separate vocals from background tracks

Configuration:

• CambSettings — Pydantic-based settings with environment variable support (CAMB_API_KEY)
• CambTextToAudioExecutionSettings / CambAudioToTextExecutionSettings — Request configuration

Other:

• Optional dependency camb-sdk added to pyproject.toml
• 40 unit tests covering all services, plugin, settings, and execution settings
• Sample script (camb_all_features.py) demonstrating all 8 features

Test Plan

• uv run pre-commit run -a — all hooks pass (ruff, bandit, pyupgrade, etc.)
• uv run mypy --config-file mypy.ini semantic_kernel/connectors/ai/camb/ — no issues in 8 source files
• uv run pytest tests/unit/connectors/ai/camb/ -v — 40/40 tests pass