search

adampower48 / AI-City-Anomaly-Detection

My implementation of the vehicle anomaly detection from https://github.com/ShuaiBai623/AI-City-Anomaly-Detection
10 stars 4 forks source link

Can you provide a fixed model to run this code? #1

Open guoguoguilai opened 4 years ago

guoguoguilai commented 4 years ago

I have tried several models downloaded from modelZoo，but they all got "The model and loaded state dict do not match exactly." Dose it matter?

adampower48 commented 4 years ago

Hi, sorry for the late reply. The version of mmdetection I used was 0.6.0+084a389. In newer versions the config file structure has been changed, so my config files will no longer work. Make sure the config files you are using are for the same version of mmdetection, and that the model you download also match.

For example, The config file: https://github.com/open-mmlab/mmdetection/blob/master/configs/faster_rcnn_r50_fpn_1x.py Model file: https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/faster_rcnn_r50_fpn_1x_20181010-3d1b3351.pth Should work together.

guoguoguilai commented 4 years ago

I feel sorry to disturb you again. When I choose Sequential by processing steps ,it throws "TypeError: 'DataFrameGroupBy' object is not an iterator". File "test.py", line 76, in <module> anomaly_results_dir, bg_interval=4, crop_min_obj_size=8) File "/disk/peizhiguo/Anomaly/AI-City-Anomaly-Detection/pipeline.py", line 277, in process_folder detector_verbose_interval, verbose) File "/disk/peizhiguo/Anomaly/AI-City-Anomaly-Detection/pipeline.py", line 124, in full_run_single similarity_thresh, suspicious_duration_thresh, verbose) File "/disk/peizhiguo/Anomaly/AI-City-Anomaly-Detection/anomaly.py", line 85, in get_anomalies_preprocessed verbose) File "/disk/peizhiguo/Anomaly/AI-City-Anomaly-Detection/anomaly.py", line 186, in get_anomalies_sequential static_results = static_results_dict[frame] File "/disk/peizhiguo/Anomaly/AI-City-Anomaly-Detection/utils.py", line 208, in __getitem__ frame, results = next(self.results_gen) TypeError: 'DataFrameGroupBy' object is not an iterator I guess the reason is the version of Pandas. What's the version of your pandas?