[AZ-180] Refactor detection event handling and improve SSE support

- Updated the detection image endpoint to require a channel ID for event streaming.
- Introduced a new endpoint for streaming detection events, allowing clients to receive real-time updates.
- Enhanced the internal buffering mechanism for detection events to manage multiple channels.
- Refactored the inference module to support the new event handling structure.

Made-with: Cursor

e2e/tests/test_performance.py

@@ -1,5 +1,4 @@
 import json
-import time
 
 import pytest
 
@@ -19,20 +18,13 @@ def _percentile_ms(sorted_ms, p):
 
 @pytest.mark.timeout(60)
 def test_nft_perf_01_single_image_latency_p95(
-    warm_engine, http_client, image_small, auth_headers
+    warm_engine, image_detect, image_small
 ):
     times_ms = []
     for _ in range(10):
-        t0 = time.perf_counter()
-        r = http_client.post(
-            "/detect/image",
-            files={"file": ("img.jpg", image_small, "image/jpeg")},
-            headers=auth_headers,
-            timeout=8,
-        )
-        elapsed_ms = (time.perf_counter() - t0) * 1000.0
-        assert r.status_code == 200
+        _, elapsed_ms = image_detect(image_small, "img.jpg", timeout=8)
         times_ms.append(elapsed_ms)
+
     sorted_ms = sorted(times_ms)
     p50 = _percentile_ms(sorted_ms, 50)
     p95 = _percentile_ms(sorted_ms, 95)
@@ -47,34 +39,16 @@ def test_nft_perf_01_single_image_latency_p95(
 
 @pytest.mark.timeout(60)
 def test_nft_perf_03_tiling_overhead_large_image(
-    warm_engine, http_client, image_small, image_large, auth_headers
+    warm_engine, image_detect, image_small, image_large
 ):
-    t_small = time.perf_counter()
-    r_small = http_client.post(
-        "/detect/image",
-        files={"file": ("small.jpg", image_small, "image/jpeg")},
-        headers=auth_headers,
-        timeout=8,
-    )
-    small_ms = (time.perf_counter() - t_small) * 1000.0
-    assert r_small.status_code == 200
-    config = json.dumps(
-        {"altitude": 400, "focal_length": 24, "sensor_width": 23.5}
-    )
-    t_large = time.perf_counter()
-    r_large = http_client.post(
-        "/detect/image",
-        files={"file": ("large.jpg", image_large, "image/jpeg")},
-        data={"config": config},
-        headers=auth_headers,
+    _, small_ms = image_detect(image_small, "small.jpg", timeout=8)
+    _, large_ms = image_detect(
+        image_large, "large.jpg",
+        config=json.dumps({"altitude": 400, "focal_length": 24, "sensor_width": 23.5}),
         timeout=20,
     )
-    large_ms = (time.perf_counter() - t_large) * 1000.0
-    assert r_large.status_code == 200
     assert large_ms < 30_000.0
     print(
         f"nft_perf_03_csv,baseline_small_ms,{small_ms:.2f},large_ms,{large_ms:.2f}"
     )
     assert large_ms > small_ms - 500.0
-
-