speaker-detector 0.1.6__py3-none-any.whl → 0.1.7__py3-none-any.whl

speaker_detector/constants.py

@@ -0,0 +1,10 @@
+# speaker_detector/constants.py
+
+BACKEND_VERSION = "0.1.7"
+
+# API base paths (optional, for future centralization)
+API_PREFIX = "/api"
+
+# Thresholds and default settings
+DEFAULT_CONFIDENCE_THRESHOLD = 0.75
+DEFAULT_INTERVAL_MS = 3000

speaker_detector/core.py

@@ -114,3 +114,5 @@ def get_speakers_needing_rebuild() -> list[str]:
 
 # ── ALIAS FOR COMPATIBILITY ──────────────────────────────────────────────────
 rebuild_embeddings_for_speaker = rebuild_embedding
+
+

speaker_detector/server.py

@@ -1,26 +1,36 @@
 # ── Core Imports ─────────────────────────────────────────────
 import os, signal, time
-from flask import Flask, request, send_from_directory, send_file
+from flask import Flask, request, send_from_directory, send_file, jsonify
 from flask_cors import CORS
 from pathlib import Path
 
 # ── Internal Modules ─────────────────────────────────────────
-from speaker_detector.state import stop_event
 from speaker_detector.utils.paths import STATIC_DIR, INDEX_HTML, COMPONENTS_DIR
+from speaker_detector.speaker_state import LISTENING_MODE, start_detection_loop, stop_detection_loop, stop_event
+from speaker_detector.constants import BACKEND_VERSION
+
+
 
 # ── App Setup ────────────────────────────────────────────────
 app = Flask(__name__, static_folder=str(STATIC_DIR))
-CORS(app)
+CORS(app, resources={r"/api/*": {"origins": "*"}}, supports_credentials=True)
 
 
 # ── Routes ──────────────────────────────────────────────────
-@app.after_request
-def apply_csp(response):
-    response.headers["Content-Security-Policy"] = (
-        "default-src 'self'; script-src 'self'; style-src 'self'; object-src 'none';"
-    )
+@app.route("/api/<path:dummy>", methods=["OPTIONS"])
+def cors_preflight(dummy):
+    response = jsonify({"ok": True})
+    response.headers["Access-Control-Allow-Origin"] = "*"
+    response.headers["Access-Control-Allow-Headers"] = "Content-Type,Authorization"
+    response.headers["Access-Control-Allow-Methods"] = "GET,POST,OPTIONS"
     return response
 
+
+
+@app.route("/api/version")
+def get_version():
+    return jsonify({"version": BACKEND_VERSION})
+
 @app.route("/")
 def serve_index():
     return send_file(INDEX_HTML)
@@ -47,22 +57,25 @@ def not_found(e):
 
 # ── Route Registrations ─────────────────────────────────────
 from speaker_detector.routes.index_routes import index_bp
-from speaker_detector.routes.settings_routes import settings_bp
+from speaker_detector.routes.listening_mode_routes import listening_bp
 from speaker_detector.routes.speaker_routes import speakers_bp
 from speaker_detector.routes.background_routes import background_bp
 from speaker_detector.routes.rebuild_routes import rebuild_bp
 from speaker_detector.routes.identify_routes import identify_bp
 from speaker_detector.routes.recordings_routes import recordings_bp
 from speaker_detector.routes.meetings_routes import meetings_bp
+from speaker_detector.routes.correction_routes import correction_bp
+
 
 app.register_blueprint(index_bp)
-app.register_blueprint(settings_bp)
+app.register_blueprint(listening_bp)
 app.register_blueprint(speakers_bp)
 app.register_blueprint(background_bp)
 app.register_blueprint(rebuild_bp)
 app.register_blueprint(identify_bp)
 app.register_blueprint(recordings_bp)
 app.register_blueprint(meetings_bp)
+app.register_blueprint(correction_bp)
 
 # ── Interrupt Handler ───────────────────────────────────────
 def handle_interrupt(sig, frame):

speaker_detector/speaker_state.py

@@ -0,0 +1,103 @@
+# speaker_detector/speaker_state.py
+
+import threading
+import tempfile
+import time
+import sounddevice as sd
+import soundfile as sf
+from datetime import datetime
+
+from speaker_detector.constants import DEFAULT_CONFIDENCE_THRESHOLD, DEFAULT_INTERVAL_MS
+from speaker_detector.core import identify_speaker
+
+# ── Shared Speaker Detection State ─────────────────────────────
+
+current_speaker_state = {
+    "speaker": None,
+    "confidence": None,
+    "is_speaking": False,
+}
+
+def get_current_speaker():
+    return current_speaker_state
+
+LISTENING_MODE = {"mode": "off"}  # Options: "off", "single", "multi"
+DETECTION_INTERVAL_MS = DEFAULT_INTERVAL_MS
+DETECTION_THRESHOLD = DEFAULT_CONFIDENCE_THRESHOLD
+
+MIC_AVAILABLE = True
+stop_event = threading.Event()
+detection_thread = None
+
+# ── Background Detection Loop ─────────────────────────────
+
+def detection_loop():
+    global MIC_AVAILABLE
+
+    samplerate = 16000
+    duration = 2  # seconds
+
+    while not stop_event.is_set():
+        try:
+            audio = sd.rec(int(duration * samplerate), samplerate=samplerate, channels=1, dtype="int16")
+            sd.wait()
+
+            with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as tmp:
+                sf.write(tmp.name, audio, samplerate)
+                MIC_AVAILABLE = True
+
+                speaker, conf = identify_speaker(tmp.name, threshold=DETECTION_THRESHOLD)
+                current_speaker_state["speaker"] = speaker
+                current_speaker_state["confidence"] = conf
+                current_speaker_state["is_speaking"] = speaker != "unknown" and conf >= DETECTION_THRESHOLD
+
+                print(f"{datetime.now().strftime('%H:%M:%S')} 🧠 Detected: {speaker} ({conf:.2f})")
+
+        except Exception as e:
+            print(f"❌ Detection loop error: {e}")
+            current_speaker_state["speaker"] = None
+            current_speaker_state["confidence"] = None
+            current_speaker_state["is_speaking"] = False
+            if isinstance(e, sd.PortAudioError):
+                MIC_AVAILABLE = False
+
+        time.sleep(DETECTION_INTERVAL_MS / 1000.0)
+
+# ── Lifecycle Control ─────────────────────────────────────
+
+def start_detection_loop():
+    global detection_thread
+    if detection_thread and detection_thread.is_alive():
+        return
+    print("🔁 Starting detection loop...")
+    stop_event.clear()
+    detection_thread = threading.Thread(target=detection_loop, daemon=True)
+    detection_thread.start()
+
+def stop_detection_loop():
+    if detection_thread and detection_thread.is_alive():
+        print("⏹️ Stopping detection loop...")
+        stop_event.set()
+
+def get_active_speaker():
+    if LISTENING_MODE["mode"] == "off":
+        return {
+            "speaker": None,
+            "confidence": None,
+            "is_speaking": False,
+            "status": "disabled"
+        }
+    if not MIC_AVAILABLE:
+        return {
+            "speaker": None,
+            "confidence": None,
+            "is_speaking": False,
+            "status": "mic unavailable"
+        }
+
+    return {
+        "speaker": current_speaker_state.get("speaker"),
+        "confidence": current_speaker_state.get("confidence"),
+        "is_speaking": current_speaker_state.get("is_speaking", False),
+        "status": "listening"
+    }

{speaker_detector-0.1.6.dist-info → speaker_detector-0.1.7.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: speaker-detector
-Version: 0.1.6
+Version: 0.1.7
 Summary: A CLI + Web tool for speaker enrollment and identification using SpeechBrain.
 Author-email: Lara Whybrow <lara.whybrow@gmail.com>
 License: MIT
@@ -29,7 +29,7 @@ Requires-Dist: sounddevice
 Requires-Dist: soundfile
 Requires-Dist: pydub
 
-Note: Still in development, as I am configuring the system for the most performant approach. Feel free to jump on the project with me. 
+23/07/2025 - Lara Whybrow, Creator - Works as far as I can see, but now we have a robust set of tools for training and managing the data and can focus now on refinement and bug fixing. 
 
 # speaker-detector 🎙️
 

{speaker_detector-0.1.6.dist-info → speaker_detector-0.1.7.dist-info}/RECORD

@@ -1,9 +1,10 @@
 speaker_detector/__main__.py,sha256=EClCwCzb6h6YBpt0hrnG4h0mlNhNePyg_xBNNSVm1os,65
 speaker_detector/cli.py,sha256=TxJhu3Pjhg41tkcu--aLtn0vZwBYyoVEef10zqSBzig,2619
-speaker_detector/core.py,sha256=zct1lNfeGB6Y7WB1rqNVPKyMKXu2g0KCuMkpAiyTLi0,5267
+speaker_detector/constants.py,sha256=DcD5Ys45loHOPmC1ORqXy2de4dYu9WsMn87FqUx1PDc,233
+speaker_detector/core.py,sha256=HiMaCiIhlm6gteLxPjciyRtbKWpuB-R6F3LTCtsV-rw,5269
 speaker_detector/server copy.py,sha256=A1WplNK8yGe9AnEjrSRqHO-uJJsMwIlvEzDhgu72XyY,10723
-speaker_detector/server.py,sha256=6AkZ2drrE-jfV3M-4jdzobontsVxrGSrmQPykpSWyeo,3516
-speaker_detector/state.py,sha256=ikcTNsjBzbFNMWfpfDaWkWMOb01Y9wlzvCAu7ifQ30w,2761
+speaker_detector/server.py,sha256=pxEx0yty0C5tlEMDr_G_zAkM4af1WYuTJzBg0bhElRc,4075
+speaker_detector/speaker_state.py,sha256=v_CfuRFjfSYCOb8VKsnmkJPbAX5AuH8xG2O7XTAywu8,3470
 speaker_detector/model/ECAPA_TDNN.py,sha256=KB5T-ye4c9ZWgTgn_SMH-T_-qYSEHQJJtf3xHjsfNPk,19024
 speaker_detector/model/classifier.ckpt,sha256=_Z42NP5ovQpCfJXjVMDGdzdPYrP0NORbeFmZUNhg1TU,5534328
 speaker_detector/model/embedding_model.ckpt,sha256=BXXLZIRea5oQ25vLdNWsMrMmuNyQNSZx00Xi7j0BJqI,83316686
@@ -18,8 +19,8 @@ speaker_detector/web/static/scripts/loader copy.js,sha256=BwhTS_ulxb62cwF6qAk1ng
 speaker_detector/web/static/scripts/loader.js,sha256=OWgmKfZ0E7bKaVPR47Q5aA-JKFeD3741K_wS_lqphz4,503
 speaker_detector/web/static/scripts/script copy.js,sha256=LLcKKjTjXEy9yj5e4gQhgIsItRLYAJe_9V7mppyM8Bc,31494
 speaker_detector/web/static/scripts/script.js,sha256=UyHWk1HrkWW6ZyxQYRLOPbtCMLHcBbNudMfTJ9k3neA,1023
-speaker_detector-0.1.6.dist-info/METADATA,sha256=WSOiZtAgOwze7mJFPETyqg7VtTOOHp4nQewa-GHUAlc,5414
-speaker_detector-0.1.6.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-speaker_detector-0.1.6.dist-info/entry_points.txt,sha256=2B30ee2cTyeeA49x_TBURl53bDRiLWGK3NWhb9rlK3s,63
-speaker_detector-0.1.6.dist-info/top_level.txt,sha256=PJ5rfvd3GAbzMbc7-Fwhtufjf6HxzzTiiHociOy7RiM,17
-speaker_detector-0.1.6.dist-info/RECORD,,
+speaker_detector-0.1.7.dist-info/METADATA,sha256=A7ixcI1t6vzezax94t5KWpwLEed8GB6Jl_hDz8RVwl8,5464
+speaker_detector-0.1.7.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+speaker_detector-0.1.7.dist-info/entry_points.txt,sha256=2B30ee2cTyeeA49x_TBURl53bDRiLWGK3NWhb9rlK3s,63
+speaker_detector-0.1.7.dist-info/top_level.txt,sha256=PJ5rfvd3GAbzMbc7-Fwhtufjf6HxzzTiiHociOy7RiM,17
+speaker_detector-0.1.7.dist-info/RECORD,,

speaker_detector/state.py

@@ -1,69 +0,0 @@
-# speaker_detector/state.py
-
-import threading
-import tempfile
-import time
-import sounddevice as sd
-import soundfile as sf
-from datetime import datetime
-from pathlib import Path
-
-from speaker_detector.core import identify_speaker  # ✅ safe import — no circular loop
-
-# ── Global State ─────────────────────────────────────────────
-current_speaker = {"speaker": None, "confidence": None}
-LISTENING_MODE = {"mode": "off"}  # Values: "off", "single", "multi"
-DETECTION_INTERVAL_MS = 3000
-DETECTION_THRESHOLD = 0.75
-
-MIC_AVAILABLE = True
-stop_event = threading.Event()  # ✅ defined here, no self-import
-detection_thread = None
-
-# ── Background Detection Loop ────────────────────────────────
-def detection_loop():
-    global MIC_AVAILABLE
-
-    samplerate = 16000
-    duration = 2
-
-    while not stop_event.is_set():
-        try:
-            audio = sd.rec(int(duration * samplerate), samplerate=samplerate, channels=1, dtype="int16")
-            sd.wait()
-
-            with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as tmp:
-                sf.write(tmp.name, audio, samplerate)
-                MIC_AVAILABLE = True
-                speaker, conf = identify_speaker(tmp.name, threshold=DETECTION_THRESHOLD)
-                current_speaker.update(speaker=speaker, confidence=conf)
-                print(f"{datetime.now().strftime('%H:%M:%S')} 🧠 Detected: {speaker} ({conf:.2f})")
-        except Exception as e:
-            print(f"❌ Detection loop error: {e}")
-            current_speaker.update(speaker=None, confidence=None)
-            if isinstance(e, sd.PortAudioError):
-                MIC_AVAILABLE = False
-
-        time.sleep(DETECTION_INTERVAL_MS / 1000.0)
-
-# ── Control Functions ────────────────────────────────────────
-def start_detection_loop():
-    global detection_thread
-    if detection_thread and detection_thread.is_alive():
-        return
-    print("🔁 Starting detection loop...")
-    stop_event.clear()
-    detection_thread = threading.Thread(target=detection_loop, daemon=True)
-    detection_thread.start()
-
-def stop_detection_loop():
-    if detection_thread and detection_thread.is_alive():
-        print("⏹️ Stopping detection loop...")
-        stop_event.set()
-
-def get_active_speaker():
-    if LISTENING_MODE["mode"] == "off":
-        return {"speaker": None, "confidence": None, "status": "disabled"}
-    if not MIC_AVAILABLE:
-        return {"speaker": None, "confidence": None, "status": "mic unavailable"}
-    return {**current_speaker, "status": "listening"}