diff --git a/SyncNetInstance.py b/SyncNetInstance.py
index 497d44f..e446be0 100644
--- a/SyncNetInstance.py
+++ b/SyncNetInstance.py
@@ -67,8 +67,15 @@ def evaluate(self, opt, videofile):
         flist = glob.glob(os.path.join(opt.tmp_dir,opt.reference,'*.jpg'))
         flist.sort()
 
+        _TARGET_SIZE = (224, 224)
         for fname in flist:
-            images.append(cv2.imread(fname))
+            img = cv2.imread(fname)
+            h, w = img.shape[:2]
+            if (w, h) != _TARGET_SIZE:
+                print("WARNING: Resizing %s from %s to %s" % (fname, (w, h), _TARGET_SIZE))
+                interp = cv2.INTER_AREA if (w > _TARGET_SIZE[0] or h > _TARGET_SIZE[1]) else cv2.INTER_LINEAR
+                img = cv2.resize(img, _TARGET_SIZE, interpolation=interp)
+            images.append(img)
 
         im = numpy.stack(images,axis=3)
         im = numpy.expand_dims(im,axis=0)
diff --git a/detectors/s3fd/box_utils.py b/detectors/s3fd/box_utils.py
index 0779bcd..1bf4be2 100644
--- a/detectors/s3fd/box_utils.py
+++ b/detectors/s3fd/box_utils.py
@@ -35,7 +35,7 @@ def nms_(dets, thresh):
         inds = np.where(ovr <= thresh)[0]
         order = order[inds + 1]
 
-    return np.array(keep).astype(np.int)
+    return np.array(keep).astype(int)
 
 
 def decode(loc, priors, variances):
diff --git a/requirements.txt b/requirements.txt
index 8919740..5614f8f 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -2,6 +2,6 @@ torch>=1.4.0
 torchvision>=0.5.0
 numpy>=1.18.1
 scipy>=1.2.1
-scenedetect==0.5.1
-opencv-contrib-python
+scenedetect
+opencv-python
 python_speech_features
diff --git a/run_visualise.py b/run_visualise.py
index 85d8925..59d3e0b 100644
--- a/run_visualise.py
+++ b/run_visualise.py
@@ -69,7 +69,7 @@
 
 	for face in faces[fidx]:
 
-		clr = max(min(face['conf']*25,255),0)
+		clr = int(max(min(face['conf']*25,255),0))
 
 		cv2.rectangle(image,(int(face['x']-face['s']),int(face['y']-face['s'])),(int(face['x']+face['s']),int(face['y']+face['s'])),(0,clr,255-clr),3)
 		cv2.putText(image,'Track %d, Conf %.3f'%(face['track'],face['conf']), (int(face['x']-face['s']),int(face['y']-face['s'])),cv2.FONT_HERSHEY_SIMPLEX,0.5,(255,255,255),2)