global_and_local: debug

configs/local/inference_config.yaml

@@ -6,67 +6,71 @@ runner:
     cuda_visible_devices: "0,1,2,3,4,5,6,7"
     
   experiment:
-    name: overfit_ab_global_only
+    name: w_gf_wo_lf_full
     root_dir: "experiments"
-    epoch: -1 # -1 stands for last epoch
+    epoch: 1 # -1 stands for last epoch
 
   test:
     dataset_list:
       - OmniObject3d_train
 
-  blender_script_path: "/data/hofee/project/nbv_rec/blender/data_renderer.py"
+  blender_script_path: "/media/hofee/data/project/python/nbv_reconstruction/blender/data_renderer.py"
-  output_dir: "/data/hofee/data/inference_global_full_on_testset"
+  output_dir: "/media/hofee/data/project/python/nbv_reconstruction/nbv_reconstruction/test/inference_global_full_on_testset"
-  pipeline: nbv_reconstruction_pipeline
+  pipeline: nbv_reconstruction_global_pts_pipeline
-  voxel_size: 0.003
   
 dataset:
   OmniObject3d_train:
-    root_dir: "/data/hofee/data/new_full_data"
+    root_dir: "/media/hofee/repository/nbv_reconstruction_data_512"
-    model_dir: "/data/hofee/data/scaled_object_meshes"
+    model_dir: "/media/hofee/data/data/scaled_object_meshes"
-    source: seq_reconstruction_dataset
+    source: seq_nbv_reconstruction_dataset
-    split_file: "/data/hofee/data/sample.txt"
+    split_file: "/media/hofee/data/project/python/nbv_reconstruction/nbv_reconstruction/test/test_set_list.txt"
     type: test
     filter_degree: 75
     ratio: 1
     batch_size: 1
     num_workers: 12
-    pts_num: 8192
+    pts_num: 4096
-    load_from_preprocess: True
+    load_from_preprocess: False
-
-  OmniObject3d_test:
-    root_dir: "/data/hofee/data/new_full_data"
-    model_dir: "/data/hofee/data/scaled_object_meshes"
-    source: seq_reconstruction_dataset
-    split_file: "/data/hofee/data/sample.txt"
-    type: test
-    filter_degree: 75
-    eval_list:
-      - pose_diff
-      - coverage_rate_increase
-    ratio: 0.1
-    batch_size: 1
-    num_workers: 12
-    pts_num: 8192
-    load_from_preprocess: True
 
 pipeline:
-  nbv_reconstruction_pipeline:
+  nbv_reconstruction_local_pts_pipeline:
     modules:
       pts_encoder: pointnet_encoder
       seq_encoder: transformer_seq_encoder
       pose_encoder: pose_encoder
       view_finder: gf_view_finder
     eps: 1e-5
+    global_scanned_feat: False
+
+  nbv_reconstruction_global_pts_pipeline:
+    modules:
+      pts_encoder: pointnet_encoder
+      pose_seq_encoder: transformer_pose_seq_encoder
+      pose_encoder: pose_encoder
+      view_finder: gf_view_finder
+    eps: 1e-5
     global_scanned_feat: True
 
+
+    
 module:
+
   pointnet_encoder:
     in_dim: 3
     out_dim: 1024
     global_feat: True
     feature_transform: False
+
   transformer_seq_encoder:
-    embed_dim: 256
+    pts_embed_dim: 1024
+    pose_embed_dim: 256
+    num_heads: 4
+    ffn_dim: 256
+    num_layers: 3
+    output_dim: 2048
+
+  transformer_pose_seq_encoder:
+    pose_embed_dim: 256
     num_heads: 4
     ffn_dim: 256
     num_layers: 3
@@ -82,8 +86,7 @@ module:
     sample_mode: ode
     sampling_steps: 500
     sde_mode: ve
+
   pose_encoder:
     pose_dim: 9
     out_dim: 256
-  pts_num_encoder:
-    out_dim: 64

configs/server/server_train_config.yaml

@@ -7,13 +7,13 @@ runner:
     parallel: False
     
   experiment:
-    name: train_ab_global_only
+    name: overfit_ab_global_and_local
     root_dir: "experiments"
     use_checkpoint: False
     epoch: -1 # -1 stands for last epoch
     max_epochs: 5000
     save_checkpoint_interval: 1
-    test_first: True  
+    test_first: False  
 
   train:
     optimizer:
@@ -25,53 +25,53 @@ runner:
   test:
     frequency: 3 # test frequency
     dataset_list:
-      - OmniObject3d_test
+      #- OmniObject3d_test
       - OmniObject3d_val
 
   pipeline: nbv_reconstruction_pipeline
   
 dataset:
   OmniObject3d_train:
-    root_dir: "/data/hofee/data/new_full_data"
+    root_dir: "/data/hofee/nbv_rec_part2_preprocessed"
     model_dir: "../data/scaled_object_meshes"
     source: nbv_reconstruction_dataset
-    split_file: "/data/hofee/data/new_full_data_list/OmniObject3d_train.txt"
+    split_file: "/data/hofee/data/sample.txt"
     type: train
     cache: True
     ratio: 1
-    batch_size: 80
+    batch_size: 32
-    num_workers: 128
+    num_workers: 16
     pts_num: 8192
     load_from_preprocess: True
 
   OmniObject3d_test:
-    root_dir: "/data/hofee/data/new_full_data"
+    root_dir: "/data/hofee/nbv_rec_part2_preprocessed"
     model_dir: "../data/scaled_object_meshes"
     source: nbv_reconstruction_dataset
-    split_file: "/data/hofee/data/new_full_data_list/OmniObject3d_test.txt"
+    split_file: "/data/hofee/data/sample.txt"
     type: test
     cache: True
     filter_degree: 75
     eval_list:
       - pose_diff
     ratio: 1
-    batch_size: 80
+    batch_size: 32
     num_workers: 12
     pts_num: 8192
     load_from_preprocess: True
 
   OmniObject3d_val:
-    root_dir: "/data/hofee/data/new_full_data"
+    root_dir: "/data/hofee/nbv_rec_part2_preprocessed"
     model_dir: "../data/scaled_object_meshes"
     source: nbv_reconstruction_dataset
-    split_file: "/data/hofee/data/new_full_data_list/OmniObject3d_train.txt"
+    split_file: "/data/hofee/data/sample.txt"
     type: test
     cache: True
     filter_degree: 75
     eval_list:
       - pose_diff
-    ratio: 0.1
+    ratio: 1
-    batch_size: 80
+    batch_size: 32
     num_workers: 12
     pts_num: 8192
     load_from_preprocess: True
@@ -92,21 +92,21 @@ module:
 
   pointnet_encoder:
     in_dim: 3
-    out_dim: 1024
+    out_dim: 512
     global_feat: True
     feature_transform: False
 
   transformer_seq_encoder:
-    embed_dim: 256
+    embed_dim: 768
     num_heads: 4
     ffn_dim: 256
     num_layers: 3
-    output_dim: 1024
+    output_dim: 2048
 
   gf_view_finder:
     t_feat_dim: 128
     pose_feat_dim: 256
-    main_feat_dim: 2048
+    main_feat_dim: 2560
     regression_head: Rx_Ry_and_T
     pose_mode: rot_matrix
     per_point_feature: False

core/nbv_dataset.py

@@ -34,7 +34,7 @@ class NBVReconstructionDataset(BaseDataset):
             #self.model_dir = config["model_dir"]
             self.filter_degree = config["filter_degree"]
         if self.type == namespace.Mode.TRAIN:
-            scale_ratio = 1
+            scale_ratio = 50
             self.datalist = self.datalist*scale_ratio
         if self.cache:
             expr_root = ConfigManager.get("runner", "experiment", "root_dir")
@@ -206,9 +206,11 @@ class NBVReconstructionDataset(BaseDataset):
             collate_data["combined_scanned_pts"] = torch.stack(
                 [torch.tensor(item["combined_scanned_pts"]) for item in batch]
             )
+                
             for key in batch[0].keys():
                 if key not in [
                     "scanned_pts",
+                    "scanned_pts_mask",
                     "scanned_n_to_world_pose_9d",
                     "best_to_world_pose_9d",
                     "combined_scanned_pts",

core/old_seq_dataset.py

@@ -1,154 +0,0 @@
-import numpy as np
-from PytorchBoot.dataset import BaseDataset
-import PytorchBoot.namespace as namespace
-import PytorchBoot.stereotype as stereotype
-from PytorchBoot.utils.log_util import Log
-import torch
-import os
-import sys
-sys.path.append(r"/home/data/hofee/project/nbv_rec/nbv_reconstruction")
-
-from utils.data_load import DataLoadUtil
-from utils.pose import PoseUtil
-from utils.pts import PtsUtil
-
-@stereotype.dataset("old_seq_nbv_reconstruction_dataset")
-class SeqNBVReconstructionDataset(BaseDataset):
-    def __init__(self, config):
-        super(SeqNBVReconstructionDataset, self).__init__(config)
-        self.type = config["type"]
-        if self.type != namespace.Mode.TEST:
-            Log.error("Dataset <seq_nbv_reconstruction_dataset> Only support test mode", terminate=True)
-        self.config = config
-        self.root_dir = config["root_dir"]
-        self.split_file_path = config["split_file"]
-        self.scene_name_list = self.load_scene_name_list()
-        self.datalist = self.get_datalist()
-        self.pts_num = config["pts_num"]
-        
-        self.model_dir = config["model_dir"]
-        self.filter_degree = config["filter_degree"]
-        self.load_from_preprocess = config.get("load_from_preprocess", False)
-        
-
-    def load_scene_name_list(self):
-        scene_name_list = []
-        with open(self.split_file_path, "r") as f:
-            for line in f:
-                scene_name = line.strip()
-                scene_name_list.append(scene_name)
-        return scene_name_list
-    
-    def get_datalist(self):
-        datalist = []
-        for scene_name in self.scene_name_list:
-            seq_num = DataLoadUtil.get_label_num(self.root_dir, scene_name)
-            scene_max_coverage_rate = 0
-            scene_max_cr_idx = 0
-
-            for seq_idx in range(seq_num):
-                label_path = DataLoadUtil.get_label_path(self.root_dir, scene_name, seq_idx)
-                label_data = DataLoadUtil.load_label(label_path)
-                max_coverage_rate = label_data["max_coverage_rate"]
-                if max_coverage_rate > scene_max_coverage_rate:
-                    scene_max_coverage_rate = max_coverage_rate
-                    scene_max_cr_idx = seq_idx
-        
-            label_path = DataLoadUtil.get_label_path(self.root_dir, scene_name, scene_max_cr_idx)
-            label_data = DataLoadUtil.load_label(label_path)
-            first_frame = label_data["best_sequence"][0]
-            best_seq_len = len(label_data["best_sequence"])
-            datalist.append({
-                    "scene_name": scene_name,
-                    "first_frame": first_frame,
-                    "max_coverage_rate": scene_max_coverage_rate,
-                    "best_seq_len": best_seq_len,
-                    "label_idx": scene_max_cr_idx,
-                })
-        return datalist
-
-    def __getitem__(self, index):
-        data_item_info = self.datalist[index]
-        first_frame_idx = data_item_info["first_frame"][0]
-        first_frame_coverage = data_item_info["first_frame"][1]
-        max_coverage_rate = data_item_info["max_coverage_rate"]
-        scene_name = data_item_info["scene_name"]
-        first_cam_info = DataLoadUtil.load_cam_info(DataLoadUtil.get_path(self.root_dir, scene_name, first_frame_idx), binocular=True)
-        first_view_path = DataLoadUtil.get_path(self.root_dir, scene_name, first_frame_idx)
-        first_left_cam_pose = first_cam_info["cam_to_world"]
-        first_center_cam_pose = first_cam_info["cam_to_world_O"]
-        first_target_point_cloud = DataLoadUtil.load_from_preprocessed_pts(first_view_path)
-        first_pts_num = first_target_point_cloud.shape[0]
-        first_downsampled_target_point_cloud = PtsUtil.random_downsample_point_cloud(first_target_point_cloud, self.pts_num)
-        first_to_world_rot_6d = PoseUtil.matrix_to_rotation_6d_numpy(np.asarray(first_left_cam_pose[:3,:3]))
-        first_to_world_trans = first_left_cam_pose[:3,3]
-        first_to_world_9d = np.concatenate([first_to_world_rot_6d, first_to_world_trans], axis=0)
-        diag = DataLoadUtil.get_bbox_diag(self.model_dir, scene_name)
-        voxel_threshold = diag*0.02
-        first_O_to_first_L_pose = np.dot(np.linalg.inv(first_left_cam_pose), first_center_cam_pose)
-        scene_path = os.path.join(self.root_dir, scene_name)
-        model_points_normals = DataLoadUtil.load_points_normals(self.root_dir, scene_name)
-
-        data_item = {
-            "first_pts_num": np.asarray(
-                first_pts_num, dtype=np.int32
-            ),
-            "first_pts": np.asarray([first_downsampled_target_point_cloud],dtype=np.float32),
-            "combined_scanned_pts": np.asarray(first_downsampled_target_point_cloud,dtype=np.float32),
-            "first_to_world_9d": np.asarray([first_to_world_9d],dtype=np.float32),
-            "scene_name": scene_name,
-            "max_coverage_rate": max_coverage_rate,
-            "voxel_threshold": voxel_threshold,
-            "filter_degree": self.filter_degree,
-            "O_to_L_pose": first_O_to_first_L_pose,
-            "first_frame_coverage": first_frame_coverage,
-            "scene_path": scene_path,
-            "model_points_normals": model_points_normals,
-            "best_seq_len": data_item_info["best_seq_len"],
-            "first_frame_id": first_frame_idx,
-        }
-        return data_item
-
-    def __len__(self):
-        return len(self.datalist)
-    
-    def get_collate_fn(self):
-        def collate_fn(batch):
-            collate_data = {}
-            collate_data["first_pts"] = [torch.tensor(item['first_pts']) for item in batch]
-            collate_data["first_to_world_9d"] = [torch.tensor(item['first_to_world_9d']) for item in batch]
-            collate_data["combined_scanned_pts"] = torch.stack([torch.tensor(item['combined_scanned_pts']) for item in batch])
-            for key in batch[0].keys():
-                if key not in ["first_pts", "first_to_world_9d", "combined_scanned_pts"]:
-                    collate_data[key] = [item[key] for item in batch]
-            return collate_data
-        return collate_fn
-
-# -------------- Debug ---------------- #
-if __name__ == "__main__":
-    import torch
-    seed = 0
-    torch.manual_seed(seed)
-    np.random.seed(seed)
-    config = {
-        "root_dir": "/home/data/hofee/project/nbv_rec/data/nbv_rec_data_512_preproc_npy",
-        "split_file": "/home/data/hofee/project/nbv_rec/data/OmniObject3d_train.txt",
-        "model_dir": "/home/data/hofee/project/nbv_rec/data/scaled_object_meshes",
-        "ratio": 0.005,
-        "batch_size": 2,
-        "filter_degree": 75,
-        "num_workers": 0,
-        "pts_num": 32684,
-        "type": namespace.Mode.TEST,
-        "load_from_preprocess": True
-    }
-    ds = SeqNBVReconstructionDataset(config)
-    print(len(ds))
-    #ds.__getitem__(10)
-    dl = ds.get_loader(shuffle=True)
-    for idx, data in enumerate(dl):
-        data = ds.process_batch(data, "cuda:0")
-        print(data)
-        # ------  Debug Start ------
-        import ipdb;ipdb.set_trace()
-        # ------  Debug End ------+

core/pipeline.py

@@ -29,6 +29,7 @@ class NBVReconstructionPipeline(nn.Module):
         
 
         self.eps = float(self.config["eps"])
+        self.enable_global_scanned_feat = self.config["global_scanned_feat"]
 
     def forward(self, data):
         mode = data["mode"]
@@ -54,7 +55,10 @@ class NBVReconstructionPipeline(nn.Module):
         return perturbed_x, random_t, target_score, std
 
     def forward_train(self, data):
+        start_time = time.time()
         main_feat = self.get_main_feat(data)
+        end_time = time.time()
+        print("get_main_feat time: ", end_time - start_time)
         """ get std """
         best_to_world_pose_9d_batch = data["best_to_world_pose_9d"]
         perturbed_x, random_t, target_score, std = self.pertube_data(
@@ -88,7 +92,9 @@ class NBVReconstructionPipeline(nn.Module):
         scanned_n_to_world_pose_9d_batch = data[
             "scanned_n_to_world_pose_9d"
         ]  # List(B): Tensor(S x 9)
-
+        scanned_pts_batch = data[
+            "scanned_pts"
+        ]
         device = next(self.parameters()).device
 
         embedding_list_batch = []
@@ -98,11 +104,13 @@ class NBVReconstructionPipeline(nn.Module):
             combined_scanned_pts_batch, require_per_point_feat=False
         )  # global_scanned_feat: Tensor(B x Dg)
 
-        for scanned_n_to_world_pose_9d in scanned_n_to_world_pose_9d_batch:
+        for scanned_n_to_world_pose_9d, scanned_pts in zip(scanned_n_to_world_pose_9d_batch, scanned_pts_batch):
             scanned_n_to_world_pose_9d = scanned_n_to_world_pose_9d.to(device)  # Tensor(S x 9)
+            scanned_pts = scanned_pts.to(device)  # Tensor(S x N x 3)
             pose_feat_seq = self.pose_encoder.encode_pose(scanned_n_to_world_pose_9d)  # Tensor(S x Dp) 
-            seq_embedding = pose_feat_seq
+            pts_feat_seq = self.pts_encoder.encode_points(scanned_pts, require_per_point_feat=False)  # Tensor(S x Dl)
-            embedding_list_batch.append(seq_embedding) # List(B): Tensor(S x (Dp))
+            seq_embedding = torch.cat([pose_feat_seq, pts_feat_seq], dim=-1) # Tensor(S x (Dp+Dl))
+            embedding_list_batch.append(seq_embedding) # List(B): Tensor(S x (Dp+Dl))
         
         seq_feat = self.seq_encoder.encode_sequence(embedding_list_batch) # Tensor(B x Ds)
         main_feat = torch.cat([seq_feat, global_scanned_feat], dim=-1) # Tensor(B x (Ds+Dg))

core/seq_dataset.py

@@ -1,195 +1,154 @@
-import numpy as np
+import numpy as np
-from PytorchBoot.dataset import BaseDataset
+from PytorchBoot.dataset import BaseDataset
-import PytorchBoot.namespace as namespace
+import PytorchBoot.namespace as namespace
-import PytorchBoot.stereotype as stereotype
+import PytorchBoot.stereotype as stereotype
-from PytorchBoot.config import ConfigManager
+from PytorchBoot.utils.log_util import Log
-from PytorchBoot.utils.log_util import Log
+import torch
-import torch
+import os
-import os
+import sys
-import sys
+sys.path.append(r"/home/data/hofee/project/nbv_rec/nbv_reconstruction")
-
+
-sys.path.append(r"/data/hofee/project/nbv_rec/nbv_reconstruction")
+from utils.data_load import DataLoadUtil
-
+from utils.pose import PoseUtil
-from utils.data_load import DataLoadUtil
+from utils.pts import PtsUtil
-from utils.pose import PoseUtil
+
-from utils.pts import PtsUtil
+@stereotype.dataset("seq_nbv_reconstruction_dataset")
-
+class SeqNBVReconstructionDataset(BaseDataset):
-
+    def __init__(self, config):
-@stereotype.dataset("seq_reconstruction_dataset")
+        super(SeqNBVReconstructionDataset, self).__init__(config)
-class SeqReconstructionDataset(BaseDataset):
+        self.type = config["type"]
-    def __init__(self, config):
+        if self.type != namespace.Mode.TEST:
-        super(SeqReconstructionDataset, self).__init__(config)
+            Log.error("Dataset <seq_nbv_reconstruction_dataset> Only support test mode", terminate=True)
-        self.config = config
+        self.config = config
-        self.root_dir = config["root_dir"]
+        self.root_dir = config["root_dir"]
-        self.split_file_path = config["split_file"]
+        self.split_file_path = config["split_file"]
-        self.scene_name_list = self.load_scene_name_list()
+        self.scene_name_list = self.load_scene_name_list()
-        self.datalist = self.get_datalist()
+        self.datalist = self.get_datalist()
-
+        self.pts_num = config["pts_num"]
-        self.pts_num = config["pts_num"]
+        
-        self.type = config["type"]
+        self.model_dir = config["model_dir"]
-        self.cache = config.get("cache")
+        self.filter_degree = config["filter_degree"]
-        self.load_from_preprocess = config.get("load_from_preprocess", False)
+        self.load_from_preprocess = config.get("load_from_preprocess", False)
-
+        
-        if self.type == namespace.Mode.TEST:
+
-            #self.model_dir = config["model_dir"]
+    def load_scene_name_list(self):
-            self.filter_degree = config["filter_degree"]
+        scene_name_list = []
-        if self.type == namespace.Mode.TRAIN:
+        with open(self.split_file_path, "r") as f:
-            scale_ratio = 1
+            for line in f:
-            self.datalist = self.datalist*scale_ratio
+                scene_name = line.strip()
-        if self.cache:
+                scene_name_list.append(scene_name)
-            expr_root = ConfigManager.get("runner", "experiment", "root_dir")
+        return scene_name_list
-            expr_name = ConfigManager.get("runner", "experiment", "name")
+    
-            self.cache_dir = os.path.join(expr_root, expr_name, "cache")
+    def get_datalist(self):
-            # self.preprocess_cache()
+        datalist = []
-
+        for scene_name in self.scene_name_list:
-    def load_scene_name_list(self):
+            seq_num = DataLoadUtil.get_label_num(self.root_dir, scene_name)
-        scene_name_list = []
+            scene_max_coverage_rate = 0
-        with open(self.split_file_path, "r") as f:
+            scene_max_cr_idx = 0
-            for line in f:
+
-                scene_name = line.strip()
+            for seq_idx in range(seq_num):
-                scene_name_list.append(scene_name)
+                label_path = DataLoadUtil.get_label_path(self.root_dir, scene_name, seq_idx)
-        return scene_name_list
+                label_data = DataLoadUtil.load_label(label_path)
-
+                max_coverage_rate = label_data["max_coverage_rate"]
-    def get_datalist(self):
+                if max_coverage_rate > scene_max_coverage_rate:
-        datalist = []
+                    scene_max_coverage_rate = max_coverage_rate
-        for scene_name in self.scene_name_list:
+                    scene_max_cr_idx = seq_idx
-            seq_num = DataLoadUtil.get_label_num(self.root_dir, scene_name)
+        
-            scene_max_coverage_rate = 0
+            label_path = DataLoadUtil.get_label_path(self.root_dir, scene_name, scene_max_cr_idx)
-            max_coverage_rate_list = []
+            label_data = DataLoadUtil.load_label(label_path)
-            scene_max_cr_idx = 0
+            first_frame = label_data["best_sequence"][0]
-            for seq_idx in range(seq_num):
+            best_seq_len = len(label_data["best_sequence"])
-                label_path = DataLoadUtil.get_label_path(
+            datalist.append({
-                    self.root_dir, scene_name, seq_idx
+                    "scene_name": scene_name,
-                )
+                    "first_frame": first_frame,
-                label_data = DataLoadUtil.load_label(label_path)
+                    "max_coverage_rate": scene_max_coverage_rate,
-                max_coverage_rate = label_data["max_coverage_rate"]
+                    "best_seq_len": best_seq_len,
-                if max_coverage_rate > scene_max_coverage_rate:
+                    "label_idx": scene_max_cr_idx,
-                    scene_max_coverage_rate = max_coverage_rate
+                })
-                    scene_max_cr_idx = seq_idx
+        return datalist
-                max_coverage_rate_list.append(max_coverage_rate)
+
-            best_label_path = DataLoadUtil.get_label_path(self.root_dir, scene_name, scene_max_cr_idx)
+    def __getitem__(self, index):
-            best_label_data = DataLoadUtil.load_label(best_label_path)
+        data_item_info = self.datalist[index]
-            first_frame = best_label_data["best_sequence"][0]
+        first_frame_idx = data_item_info["first_frame"][0]
-            best_seq_len = len(best_label_data["best_sequence"])
+        first_frame_coverage = data_item_info["first_frame"][1]
-            datalist.append({
+        max_coverage_rate = data_item_info["max_coverage_rate"]
-                "scene_name": scene_name,
+        scene_name = data_item_info["scene_name"]
-                "first_frame": first_frame,
+        first_cam_info = DataLoadUtil.load_cam_info(DataLoadUtil.get_path(self.root_dir, scene_name, first_frame_idx), binocular=True)
-                "best_seq_len": best_seq_len,
+        first_view_path = DataLoadUtil.get_path(self.root_dir, scene_name, first_frame_idx)
-                "max_coverage_rate": scene_max_coverage_rate,
+        first_left_cam_pose = first_cam_info["cam_to_world"]
-                "label_idx": scene_max_cr_idx,
+        first_center_cam_pose = first_cam_info["cam_to_world_O"]
-            })  
+        first_target_point_cloud = DataLoadUtil.load_from_preprocessed_pts(first_view_path)
-        return datalist
+        first_pts_num = first_target_point_cloud.shape[0]
-
+        first_downsampled_target_point_cloud = PtsUtil.random_downsample_point_cloud(first_target_point_cloud, self.pts_num)
-    def preprocess_cache(self):
+        first_to_world_rot_6d = PoseUtil.matrix_to_rotation_6d_numpy(np.asarray(first_left_cam_pose[:3,:3]))
-        Log.info("preprocessing cache...")
+        first_to_world_trans = first_left_cam_pose[:3,3]
-        for item_idx in range(len(self.datalist)):
+        first_to_world_9d = np.concatenate([first_to_world_rot_6d, first_to_world_trans], axis=0)
-            self.__getitem__(item_idx)
+        diag = DataLoadUtil.get_bbox_diag(self.model_dir, scene_name)
-        Log.success("finish preprocessing cache.")
+        voxel_threshold = diag*0.02
-
+        first_O_to_first_L_pose = np.dot(np.linalg.inv(first_left_cam_pose), first_center_cam_pose)
-    def load_from_cache(self, scene_name, curr_frame_idx):
+        scene_path = os.path.join(self.root_dir, scene_name)
-        cache_name = f"{scene_name}_{curr_frame_idx}.txt"
+        model_points_normals = DataLoadUtil.load_points_normals(self.root_dir, scene_name)
-        cache_path = os.path.join(self.cache_dir, cache_name)
+
-        if os.path.exists(cache_path):
+        data_item = {
-            data = np.loadtxt(cache_path)
+            "first_pts_num": np.asarray(
-            return data
+                first_pts_num, dtype=np.int32
-        else:
+            ),
-            return None
+            "first_pts": np.asarray([first_downsampled_target_point_cloud],dtype=np.float32),
-
+            "combined_scanned_pts": np.asarray(first_downsampled_target_point_cloud,dtype=np.float32),
-    def save_to_cache(self, scene_name, curr_frame_idx, data):
+            "first_to_world_9d": np.asarray([first_to_world_9d],dtype=np.float32),
-        cache_name = f"{scene_name}_{curr_frame_idx}.txt"
+            "scene_name": scene_name,
-        cache_path = os.path.join(self.cache_dir, cache_name)
+            "max_coverage_rate": max_coverage_rate,
-        try:
+            "voxel_threshold": voxel_threshold,
-            np.savetxt(cache_path, data)
+            "filter_degree": self.filter_degree,
-        except Exception as e:
+            "O_to_L_pose": first_O_to_first_L_pose,
-            Log.error(f"Save cache failed: {e}")
+            "first_frame_coverage": first_frame_coverage,
-
+            "scene_path": scene_path,
-    def seq_combined_pts(self, scene, frame_idx_list):
+            "model_points_normals": model_points_normals,
-        all_combined_pts = []   
+            "best_seq_len": data_item_info["best_seq_len"],
-        for i in frame_idx_list:
+            "first_frame_id": first_frame_idx,
-            path = DataLoadUtil.get_path(self.root_dir, scene, i)
+        }
-            pts = DataLoadUtil.load_from_preprocessed_pts(path,"npy")
+        return data_item
-            if pts.shape[0] == 0:
+
-                continue
+    def __len__(self):
-            all_combined_pts.append(pts)
+        return len(self.datalist)
-        all_combined_pts = np.vstack(all_combined_pts)
+    
-        downsampled_all_pts = PtsUtil.voxel_downsample_point_cloud(all_combined_pts, 0.003)
+    def get_collate_fn(self):
-        return downsampled_all_pts
+        def collate_fn(batch):
-
+            collate_data = {}
-    def __getitem__(self, index):
+            collate_data["first_pts"] = [torch.tensor(item['first_pts']) for item in batch]
-        data_item_info = self.datalist[index]
+            collate_data["first_to_world_9d"] = [torch.tensor(item['first_to_world_9d']) for item in batch]
-        max_coverage_rate = data_item_info["max_coverage_rate"]
+            collate_data["combined_scanned_pts"] = torch.stack([torch.tensor(item['combined_scanned_pts']) for item in batch])
-        scene_name = data_item_info["scene_name"]
+            for key in batch[0].keys():
-        (
+                if key not in ["first_pts", "first_to_world_9d", "combined_scanned_pts"]:
-            scanned_views_pts,
+                    collate_data[key] = [item[key] for item in batch]
-            scanned_coverages_rate,
+            return collate_data
-            scanned_n_to_world_pose,
+        return collate_fn
-        ) = ([], [], [])
+
-        view = data_item_info["first_frame"]
+# -------------- Debug ---------------- #
-        frame_idx = view[0]
+if __name__ == "__main__":
-        coverage_rate = view[1]
+    import torch
-        view_path = DataLoadUtil.get_path(self.root_dir, scene_name, frame_idx)
+    seed = 0
-        cam_info = DataLoadUtil.load_cam_info(view_path, binocular=True)
+    torch.manual_seed(seed)
-        
+    np.random.seed(seed)
-        n_to_world_pose = cam_info["cam_to_world"]  
+    config = {
-        target_point_cloud = (
+        "root_dir": "/home/data/hofee/project/nbv_rec/data/nbv_rec_data_512_preproc_npy",
-            DataLoadUtil.load_from_preprocessed_pts(view_path)
+        "split_file": "/home/data/hofee/project/nbv_rec/data/OmniObject3d_train.txt",
-        )
+        "model_dir": "/home/data/hofee/project/nbv_rec/data/scaled_object_meshes",
-        downsampled_target_point_cloud = PtsUtil.random_downsample_point_cloud(
+        "ratio": 0.005,
-            target_point_cloud, self.pts_num
+        "batch_size": 2,
-        )
+        "filter_degree": 75,
-        scanned_views_pts.append(downsampled_target_point_cloud)
+        "num_workers": 0,
-        scanned_coverages_rate.append(coverage_rate)
+        "pts_num": 32684,
-        n_to_world_6d = PoseUtil.matrix_to_rotation_6d_numpy(
+        "type": namespace.Mode.TEST,
-            np.asarray(n_to_world_pose[:3, :3])
+        "load_from_preprocess": True
-        ) 
+    }
-        first_left_cam_pose = cam_info["cam_to_world"]
+    ds = SeqNBVReconstructionDataset(config)
-        first_center_cam_pose = cam_info["cam_to_world_O"]
+    print(len(ds))
-        first_O_to_first_L_pose = np.dot(np.linalg.inv(first_left_cam_pose), first_center_cam_pose)
+    #ds.__getitem__(10)
-        n_to_world_trans = n_to_world_pose[:3, 3]
+    dl = ds.get_loader(shuffle=True)
-        n_to_world_9d = np.concatenate([n_to_world_6d, n_to_world_trans], axis=0)
+    for idx, data in enumerate(dl):
-        scanned_n_to_world_pose.append(n_to_world_9d)
+        data = ds.process_batch(data, "cuda:0")
-
+        print(data)
-        frame_list = []
+        # ------  Debug Start ------
-        for i in range(DataLoadUtil.get_scene_seq_length(self.root_dir, scene_name)):
+        import ipdb;ipdb.set_trace()
-            frame_list.append(i)
+        # ------  Debug End ------+
-        gt_pts = self.seq_combined_pts(scene_name, frame_list)
-        data_item = {
-            "first_scanned_pts": np.asarray(scanned_views_pts, dtype=np.float32), # Ndarray(S x Nv x 3)
-            "first_scanned_coverage_rate": scanned_coverages_rate, # List(S): Float, range(0, 1)
-            "first_scanned_n_to_world_pose_9d": np.asarray(scanned_n_to_world_pose, dtype=np.float32), # Ndarray(S x 9)
-            "seq_max_coverage_rate": max_coverage_rate, # Float, range(0, 1)
-            "scene_name": scene_name, # String
-            "gt_pts": gt_pts, # Ndarray(N x 3)
-            "scene_path": os.path.join(self.root_dir, scene_name), # String
-            "O_to_L_pose": first_O_to_first_L_pose,
-        }
-
-        return data_item
-
-    def __len__(self):
-        return len(self.datalist)
-
-
-# -------------- Debug ---------------- #
-if __name__ == "__main__":
-    import torch
-
-    seed = 0
-    torch.manual_seed(seed)
-    np.random.seed(seed)
-    config = {
-        "root_dir": "/data/hofee/data/new_full_data",
-        "source": "seq_reconstruction_dataset",
-        "split_file": "/data/hofee/data/sample.txt",
-        "load_from_preprocess": True,
-        "ratio": 0.5,
-        "batch_size": 2,
-        "filter_degree": 75,
-        "num_workers": 0,
-        "pts_num": 4096,
-        "type": namespace.Mode.TRAIN,
-    }
-    ds = SeqReconstructionDataset(config)
-    print(len(ds))
-    print(ds.__getitem__(10))
-    

runners/inferencer.py

@@ -27,7 +27,6 @@ class Inferencer(Runner):
     
         self.script_path = ConfigManager.get(namespace.Stereotype.RUNNER, "blender_script_path")
         self.output_dir = ConfigManager.get(namespace.Stereotype.RUNNER, "output_dir")
-        self.voxel_size = ConfigManager.get(namespace.Stereotype.RUNNER, "voxel_size")
         ''' Pipeline '''
         self.pipeline_name = self.config[namespace.Stereotype.PIPELINE]
         self.pipeline:torch.nn.Module = ComponentFactory.create(namespace.Stereotype.PIPELINE, self.pipeline_name)
@@ -66,11 +65,16 @@ class Inferencer(Runner):
             for dataset_idx, test_set in enumerate(self.test_set_list):
                 status_manager.set_progress("inference", "inferencer", f"dataset", dataset_idx, len(self.test_set_list))
                 test_set_name = test_set.get_name()
+                test_loader = test_set.get_loader()
 
-                total=int(len(test_set))
+                if test_loader.batch_size > 1:
-                for i in range(total):
+                    Log.error("Batch size should be 1 for inference, found {} in {}".format(test_loader.batch_size, test_set_name), terminate=True)
-                    data = test_set.__getitem__(i)
+
+                total=int(len(test_loader))
+                loop = tqdm(enumerate(test_loader), total=total)
+                for i, data in loop:
                     status_manager.set_progress("inference", "inferencer", f"Batch[{test_set_name}]", i+1, total)
+                    test_set.process_batch(data, self.device)
                     output = self.predict_sequence(data)
                     self.save_inference_result(test_set_name, data["scene_name"][0], output)
                     
@@ -84,23 +88,26 @@ class Inferencer(Runner):
         ''' data for rendering '''
         scene_path = data["scene_path"][0]
         O_to_L_pose = data["O_to_L_pose"][0]
-        voxel_threshold = self.voxel_size
+        voxel_threshold = data["voxel_threshold"][0]
-        filter_degree = 75
+        filter_degree = data["filter_degree"][0]
-        down_sampled_model_pts = data["gt_pts"]
+        model_points_normals = data["model_points_normals"][0]
-        import ipdb; ipdb.set_trace()
+        model_pts = model_points_normals[:,:3]
-        first_frame_to_world_9d = data["first_scanned_n_to_world_pose_9d"][0]
+        down_sampled_model_pts = PtsUtil.voxel_downsample_point_cloud(model_pts, voxel_threshold)
-        first_frame_to_world = np.eye(4)
+        first_frame_to_world_9d = data["first_to_world_9d"][0]
-        first_frame_to_world[:3,:3] = PoseUtil.rotation_6d_to_matrix_numpy(first_frame_to_world_9d[:6])
+        first_frame_to_world = torch.eye(4, device=first_frame_to_world_9d.device)
-        first_frame_to_world[:3,3] = first_frame_to_world_9d[6:]
+        first_frame_to_world[:3,:3] = PoseUtil.rotation_6d_to_matrix_tensor_batch(first_frame_to_world_9d[:,:6])[0]
+        first_frame_to_world[:3,3] = first_frame_to_world_9d[0,6:]
+        first_frame_to_world = first_frame_to_world.to(self.device)
         
         ''' data for inference '''
         input_data = {}
-        input_data["combined_scanned_pts"] = torch.tensor(data["first_scanned_pts"][0], dtype=torch.float32).to(self.device)
+        input_data["scanned_pts"] = [data["first_pts"][0].to(self.device)]
-        input_data["scanned_n_to_world_pose_9d"] = [torch.tensor(data["first_scanned_n_to_world_pose_9d"], dtype=torch.float32).to(self.device)]
+        input_data["scanned_n_to_world_pose_9d"] = [data["first_to_world_9d"][0].to(self.device)]
         input_data["mode"] = namespace.Mode.TEST
-        input_pts_N = input_data["combined_scanned_pts"].shape[1]
+        input_data["combined_scanned_pts"] = data["combined_scanned_pts"]
+        input_pts_N = input_data["scanned_pts"][0].shape[1]
         
-        first_frame_target_pts, _ = RenderUtil.render_pts(first_frame_to_world, scene_path, self.script_path, down_sampled_model_pts, voxel_threshold=voxel_threshold, filter_degree=filter_degree, nO_to_nL_pose=O_to_L_pose)
+        first_frame_target_pts, _ = RenderUtil.render_pts(first_frame_to_world, scene_path, self.script_path, model_points_normals, voxel_threshold=voxel_threshold, filter_degree=filter_degree, nO_to_nL_pose=O_to_L_pose)
         scanned_view_pts = [first_frame_target_pts]
         last_pred_cr = self.compute_coverage_rate(scanned_view_pts, None, down_sampled_model_pts, threshold=voxel_threshold)
 

utils/render.py

@@ -10,7 +10,7 @@ from utils.pts import PtsUtil
 class RenderUtil:
 
     @staticmethod
-    def render_pts(cam_pose, scene_path, script_path, voxel_threshold=0.005, filter_degree=75, nO_to_nL_pose=None, require_full_scene=False):
+    def render_pts(cam_pose, scene_path, script_path, model_points_normals, voxel_threshold=0.005, filter_degree=75, nO_to_nL_pose=None, require_full_scene=False):
         
         nO_to_world_pose = DataLoadUtil.get_real_cam_O_from_cam_L(cam_pose, nO_to_nL_pose, scene_path=scene_path)
 
@@ -34,10 +34,10 @@ class RenderUtil:
                 return None
             path = os.path.join(temp_dir, "tmp")
             point_cloud = DataLoadUtil.get_target_point_cloud_world_from_path(path, binocular=True)
-            normals = DataLoadUtil.get_target_normals_world_from_path(path, binocular=True)
             cam_params = DataLoadUtil.load_cam_info(path, binocular=True)
             
-            filtered_point_cloud = PtsUtil.filter_points(point_cloud, normals, cam_pose=cam_params["cam_to_world"], voxel_size=voxel_threshold, theta=filter_degree)
+            ''' TODO: old code: filter_points api is changed, need to update the code '''
+            filtered_point_cloud = PtsUtil.filter_points(point_cloud, model_points_normals, cam_pose=cam_params["cam_to_world"], voxel_size=voxel_threshold, theta=filter_degree)
             full_scene_point_cloud = None
             if require_full_scene:
                 depth_L, depth_R = DataLoadUtil.load_depth(path, cam_params['near_plane'], cam_params['far_plane'], binocular=True)