display model names above top row

app.py

@@ -3,7 +3,7 @@ import subprocess
 
 import streamlit as st
 
-from utils import get_configs, get_display_names, get_path_for_viz, get_video_height
+from utils import get_configs, get_display_names, get_path_for_viz, get_video_height, get_text_str
 
 # st.header("EVREAL  - Event-based Video Reconstruction Evaluation and Analysis Library")
 #
@@ -99,12 +99,14 @@ if len(gt_viz) > 0:
     selected_models.append(ground_truth)
 
 padding = 2
+font_size = 20
 num_cols = len(selected_models)
 crop_str = "crop=trunc(iw/2)*2:trunc(ih/2)*2"
 pad_str = "pad=ceil(iw/2)*2+{}:ceil(ih/2)*2+{}:{}:{}".format(padding*2, padding*2, padding, padding)
 num_elements = num_rows * num_cols
 
 w = selected_dataset["width"]
+h = selected_dataset["height"]
 input_filter_parts = []
 xstack_input_parts = []
 layout_parts = []
@@ -133,7 +135,16 @@ for row_idx in range(num_rows):
 
         if row_heights[row_idx] == "":
             row_heights[row_idx] = "h{}".format(vid_idx)
-        input_filter_part = "[{}:v]scale={}:-1,{}[v{}]".format(vid_idx, w, pad_str, vid_idx)
+        if row_idx == 0:
+            # font_size
+            # pad=240:180+23:0:180:white
+            pad_height = font_size+padding*2
+            pad_txt_str = ",pad={}:{}:0:{}:white".format(w+padding*2, h+font_size+padding*4, pad_height)
+            text_str = get_text_str(pad_height, w, cur_model['display_name'], font_path, font_size)
+            pad_txt_str = pad_txt_str + "," + text_str
+        else:
+            pad_txt_str = ""
+        input_filter_part = "[{}:v]scale={}:-1,{}{}[v{}]".format(vid_idx, w, pad_str, pad_txt_str, vid_idx)
         input_filter_parts.append(input_filter_part)
         xstack_input_part = "[v{}]".format(vid_idx)
         xstack_input_parts.append(xstack_input_part)

utils.py

@@ -37,8 +37,10 @@ def get_path_for_viz(base_path, dataset, sequence, model, viz):
     return video_path
 
 
-def get_text_video(video_h, video_w, text, font, fontsize, fontcolor="black"):
-    pass
+def get_text_str(height, width, text, font, fontsize, fontcolor="black"):
+    return "drawtext=fontfile='{}':text='{}':" \
+           "fontsize={}:fontcolor={}:" \
+           "x=({}-text_w)/2:y=({}-text_h)/2".format(font, text, fontsize, fontcolor, width, height)
 
 
 def get_video_height(video_path):