render_flyaround bugfix

Summary: Fixes a bug which would crash render_flyaround anytime  visualize_preds_keys is adjusted

Reviewed By: shapovalov

Differential Revision: D41124462

fbshipit-source-id: 127045a91a055909f8bd56c8af81afac02c00f60

Author: davnov134

pytorch3d/implicitron/models/visualization/render_flyaround.py

@@ -10,7 +10,17 @@
 import math
 import os
 import random
-from typing import Any, Dict, List, Optional, Sequence, Tuple, TYPE_CHECKING, Union
+from typing import (
+    Any,
+    Dict,
+    Iterable,
+    List,
+    Optional,
+    Sequence,
+    Tuple,
+    TYPE_CHECKING,
+    Union,
+)
 
 import numpy as np
 import torch
@@ -180,7 +190,7 @@ def render_flyaround(
             preds.update(net_input)  # merge everything into one big dict
 
             # Render the predictions to images
-            rendered_pred = _images_from_preds(preds)
+            rendered_pred = _images_from_preds(preds, extract_keys=visualize_preds_keys)
             preds_total.append(rendered_pred)
 
             # show the preds every 5% of the export iterations
@@ -223,17 +233,20 @@ def _load_whole_dataset(
     return next(iter(load_all_dataloader))
 
 
-def _images_from_preds(preds: Dict[str, Any]) -> Dict[str, torch.Tensor]:
-    imout = {}
-    for k in (
+def _images_from_preds(
+    preds: Dict[str, Any],
+    extract_keys: Iterable[str] = (
         "image_rgb",
         "images_render",
         "fg_probability",
         "masks_render",
         "depths_render",
         "depth_map",
         "_all_source_images",
-    ):
+    ),
+) -> Dict[str, torch.Tensor]:
+    imout = {}
+    for k in extract_keys:
         if k == "_all_source_images" and "image_rgb" in preds:
             src_ims = preds["image_rgb"][1:].cpu().detach().clone()
             v = _stack_images(src_ims, None)[None]
@@ -343,6 +356,9 @@ def _generate_prediction_videos(
     # init a video writer for each predicted key
     vws = {}
     for k in predicted_keys:
+        if k not in preds[0]:
+            logger.warn(f"Cannot generate video for prediction key '{k}'")
+            continue
         cache_dir = (
             None
             if video_frames_dir is None
@@ -355,13 +371,15 @@ def _generate_prediction_videos(
         )
 
     for rendered_pred in tqdm(preds):
-        for k in predicted_keys:
+        for k in vws:
             vws[k].write_frame(
                 rendered_pred[k][0].clip(0.0, 1.0).detach().cpu().numpy(),
                 resize=resize,
             )
 
     for k in predicted_keys:
+        if k not in vws:
+            continue
         vws[k].get_video()
         logger.info(f"Generated {vws[k].out_path}.")
         if viz is not None: