diff --git a/tt-metal-yolov4/server/fast_api_yolov4.py b/tt-metal-yolov4/server/fast_api_yolov4.py
index 1e759ae8..6f26ab0e 100644
--- a/tt-metal-yolov4/server/fast_api_yolov4.py
+++ b/tt-metal-yolov4/server/fast_api_yolov4.py
@@ -4,6 +4,7 @@
 import os
 import logging
 from fastapi import FastAPI, File, HTTPException, Request, status, UploadFile
+from fastapi.responses import JSONResponse
 from functools import wraps
 from io import BytesIO
 import jwt
@@ -63,6 +64,8 @@ def load_class_names(namesfile):
     class_names = load_class_names(file_path)
 
     global model
+    global ready
+    ready = False
     if ("WH_ARCH_YAML" in os.environ) and os.environ[
         "WH_ARCH_YAML"
     ] == "wormhole_b0_80_arch_eth_dispatch.yaml":
@@ -92,6 +95,17 @@ def load_class_names(namesfile):
         ttnn.enable_program_cache(device)
         model = Yolov4Trace2CQ()
         model.initialize_yolov4_trace_2cqs_inference(device)
+    ready = True
+
+
+@app.get("/health")
+async def health_check():
+    if not ready:
+        raise HTTPException(
+            status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
+            detail="Server is not ready yet",
+        )
+    return JSONResponse(content={"message": "OK\n"}, status_code=status.HTTP_200_OK)
 
 
 @app.on_event("shutdown")
@@ -259,6 +273,7 @@ async def objdetection_v2(request: Request, file: UploadFile = File(...)):
     contents = await file.read()
     # Load and convert the image to RGB
     image = Image.open(BytesIO(contents)).convert("RGB")
+    image = image.resize((320, 320))  # Resize to target dimensions
     image = np.array(image)
     if isinstance(image, np.ndarray) and len(image.shape) == 3:  # cv2 image
         image = torch.from_numpy(image).float().div(255.0).unsqueeze(0)