this should be in a different repo

scripts/collage_comparison/vre.sh

@@ -2,4 +2,4 @@
 set -ex
 video_file=$1
 shift
-vre_gpu_parallel $video_file --gpus $@ -- --config_path /scratch/sdc/datasets/dronescapes-2024/scripts/collage_comparison/cfg.yaml -o data_${video_file} --representations semantic_mask2former_coco_47429163_0 semantic_mask2former_mapillary_49189528_0 semantic_mask2former_mapillary_49189528_1 depth_marigold "normals_svd(depth_marigold)" semantic_mask2former_swin_mapillary_converted semantic_mask2former_r50_mapillary_converted semantic_mask2former_swin_coco_converted semantic_median_expert buildings "buildings(nearby)" containing rgb safe-landing-no-sseg safe-landing-semantics sky-and-water transportation vegetation -I /export/home/proiecte/aux/mihai_cristian.pirvu/code/neo-transformers/readers/semantic_mapper.py:get_new_semantic_mapped_tasks --output_dir_exists_mode skip_computed
+vre_gpu_parallel $video_file --config_path /scratch/sdc/datasets/dronescapes-2024/scripts/collage_comparison/cfg.yaml -o data_${video_file} --representations semantic_mask2former_coco_47429163_0 semantic_mask2former_mapillary_49189528_0 semantic_mask2former_mapillary_49189528_1 depth_marigold "normals_svd(depth_marigold)" semantic_mask2former_swin_mapillary_converted semantic_mask2former_r50_mapillary_converted semantic_mask2former_swin_coco_converted semantic_median_expert buildings "buildings(nearby)" containing rgb safe-landing-no-sseg safe-landing-semantics sky-and-water transportation vegetation -I /export/home/proiecte/aux/mihai_cristian.pirvu/code/neo-transformers/readers/semantic_mapper.py:get_new_semantic_mapped_tasks --output_dir_exists_mode skip_computed

scripts/collage_comparison/wip.py

@@ -76,14 +76,13 @@ def load_model_from_path(weights_path):
     logger.info(f"Excluded (fully masked) tasks: {cfg.train.algorithm.masking.parameters.excluded_tasks}")
     return model
 
-
 def colorize_dronescapes(item: np.ndarray) -> np.ndarray:
     # colorize_semantic_segmentation
-    assert len(item.shape) == 3 and item.shape[-1] == 8, item.shape
+    assert len(item.shape) == 2, item.shape
     color_map = [[0, 255, 0], [0, 127, 0], [255, 255, 0], [255, 255, 255],
                  [255, 0, 0], [0, 0, 255], [0, 255, 255], [127, 127, 63]]
     classes_8 = ["land", "forest", "residential", "road", "little-objects", "water", "sky", "hill"]
-    return colorize_semantic_segmentation(item[None].argmax(-1), color_map=color_map, classes=classes_8)[0]
+    return colorize_semantic_segmentation(item[None], color_map=color_map, classes=classes_8)[0]
 
 @tr.no_grad
 def inference(model: LME | str, batch: dict, n_ens: int | None = None) -> np.ndarray:
@@ -111,7 +110,7 @@ def inference(model: LME | str, batch: dict, n_ens: int | None = None) -> np.nda
                 else:
                     acc_sema = (acc_sema * i + curr_sema) / (i + 1)
             item = acc_sema[0]
-    return colorize_dronescapes(item.permute(1, 2, 0).numpy())
+    return item.permute(1, 2, 0).numpy().argmax(-1).astype(np.uint8)
 
 def get_args() -> Namespace:
     parser = ArgumentParser()
@@ -179,16 +178,30 @@ def main(args: Namespace):
     fix_plot_fns_(plot_fns, task_types, stats, cfg["data"]["parameters"]["normalization"])
 
     (out_dir := Path.cwd() / f"out_{video_path.name}").mkdir(exist_ok=True)
+    [(out_dir / x).mkdir(exist_ok=True) for x in ["ens", "m2f", "distil", "collage"]]
+    
     for frame_ix in tqdm(frames):
-        if (out_file := out_dir/ f"{frame_ix}.jpg").exists():
+        if (out_file := out_dir / f"collage/{frame_ix}.jpg").exists():
             continue
         batch_m2f = reader2.collate_fn([reader2[frame_ix]])
         batch = reader.collate_fn([reader[frame_ix]])
-        m2f_img = inference("semantic_mask2former_r50_mapillary_converted", batch_m2f)
-        ens_img = inference(model_mae, batch, n_ens=30)
-        distil_img = inference(model_distil, batch)
         rgb = batch_m2f["data"]["rgb"][0].permute(1, 2, 0).numpy()
 
+        if not (pth := out_dir / f"m2f/{frame_ix}.npz").exists():
+            y = inference("semantic_mask2former_r50_mapillary_converted", batch_m2f)
+            np.savez_compressed(pth, y)
+        m2f_img = colorize_dronescapes(np.load(pth)["arr_0"])
+
+        if not (pth := out_dir / f"ens/{frame_ix}.npz").exists():
+            y = inference(model_mae, batch, n_ens=30)
+            np.savez_compressed(pth, y)
+        ens_img = colorize_dronescapes(np.load(pth)["arr_0"])
+
+        if not (pth := out_dir / f"distil/{frame_ix}.npz").exists():
+            y = inference(model_distil, batch)
+            np.savez_compressed(pth, y)
+        distil_img = colorize_dronescapes(np.load(pth)["arr_0"])
+
         titles = ["RGB", "Mask2Former (216M)", "Ensembles-30 (4M)", "Distillation (4M)"]
         collage = collage_fn([rgb, m2f_img, ens_img, distil_img], titles=titles, rows_cols=(2, 2), size_px=40)
         image_write(collage, out_file)