IndexError when reconstruct the mesh from MipNerf360 dataset

[zxy-bjtu]

When I rendered the Room in mipnerf-360 dataset, the error encountered:

# CUDA_VISIBLE_DEVICES=0 python render.py -m output/mipnerf360/room --skip_test --use_depth_filter
Looking for config file in output/mipnerf360/room/cfg_args
Config file found: output/mipnerf360/room/cfg_args
Rendering output/mipnerf360/room
multi_view_num 8 [03/09 23:48:22]
Loading trained model at iteration 30000 [03/09 23:48:22]
Reading camera 311/311 [03/09 23:48:25]
Converting point3d.bin to .ply, will happen only the first time you open the scene. [03/09 23:48:25]
xyz (100187, 3) [03/09 23:48:26]
cameras_extent 6.422796392440796 [03/09 23:48:27]
Loading Training Cameras [03/09 23:48:27]
load camera 271Loading Test Cameras [03/09 23:50:59]
load camera 38computing nearest_id [03/09 23:51:19]
extract_triangle_mesh [03/09 23:51:24]
[Open3D WARNING] Write PLY failed: mesh has 0 vertices. [03/09 23:51:24]
post processing the mesh to have 1 clusterscluster_to_kep [03/09 23:51:24]
[Open3D DEBUG] [ClusterConnectedTriangles] Compute triangle adjacency [03/09 23:51:24]
[Open3D DEBUG] [ClusterConnectedTriangles] Done computing triangle adjacency [03/09 23:51:24]
[Open3D DEBUG] [ClusterConnectedTriangles] Done clustering, #clusters=0 [03/09 23:51:24]
Traceback (most recent call last):
  File "/data1/thesis_project/PGSR/render.py", line 194, in <module>
    render_sets(model.extract(args), args.iteration, pipeline.extract(args), args.skip_train, args.skip_test, args.max_depth, args.voxel_size, args.num_cluster, args.use_depth_filter)
  File "/data1/thesis_project/PGSR/render.py", line 166, in render_sets
    mesh = post_process_mesh(mesh, num_cluster)
  File "/data1/thesis_project/PGSR/render.py", line 55, in post_process_mesh
    n_cluster = np.sort(cluster_n_triangles.copy())[-cluster_to_keep]
IndexError: index -1 is out of bounds for axis 0 with size 0

Among the 9 scenes in the mipnerf dataset, I am unable to obtain the tsdf_fusion_post.ply file only for the ​​room scene​​. How should I resolve this issue?
@danpeng2