Renamed to GraVi-T

README.md

@@ -1,9 +1,9 @@
-# Graph-LTVU
-This repository contains an open-source codebase for graph-based long-term video understanding (Graph-LTVU). It is designed to serve as a spatial-temporal graph learning framework for multiple video understanding tasks. In the current version, it supports training and evaluating one of the state-of-the-art models, [SPELL](https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136950367.pdf), for the tasks of active speaker detection and action localization.
+# GraVi-T
+This repository contains an open-source codebase for Graph-based long-term Video undersTanding (GraVi-T). It is designed to serve as a spatial-temporal graph learning framework for multiple video understanding tasks. In the current version, it supports training and evaluating one of the state-of-the-art models, [SPELL](https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136950367.pdf), for the tasks of active speaker detection and action localization.
 
 In the near future, we will release more advanced graph-based approaches for other tasks, including action segmentation and the [Ego4D challenge](https://ego4d-data.org/workshops/eccv22).
 
-![](docs/images/graphltvu_teaser.jpg?raw=true)
+![](docs/images/gravit_teaser.jpg?raw=true)
 
 ## Use Cases and Performance
 |  Model  |         Dataset         |            Task           |     validation mAP (%)     |
@@ -28,8 +28,8 @@ Alternatively, you can manually install PyYAML, pandas, and [PyG](https://www.py
 ## Installation
 After confirming the above requirements are met, run the following commands:
 ```
-https://github.com/IntelLabs/Graph-LTVU
-cd Graph-LTVU
+git clone https://github.com/IntelLabs/GraVi-T.git
+cd GraVi-T
 pip3 install -e .
 ```
 
@@ -92,7 +92,7 @@ This will print the evaluation score.
 Please refer to the instructions in [GETTING_STARTED_AL.md](docs/GETTING_STARTED_AL.md).
 
 ## Contributor
-Graph-LTVU is written and maintained by [Kyle Min](https://sites.google.com/view/kylemin)
+GraVi-T is written and maintained by [Kyle Min](https://sites.google.com/view/kylemin)
 
 ## Citation
 ECCV 2022 paper:

graphltvu/__init__.py → gravit/__init__.py

@@ -1,6 +1,6 @@
 from pkg_resources import get_distribution
 
 try:
-    __version__ = get_distribution('graphltvu').version
+    __version__ = get_distribution('gravit').version
 except:
     __version__ = '1.0.0'

graphltvu/utils/formatter.py → gravit/utils/formatter.py

@@ -16,7 +16,7 @@ def get_formatting_data_dict(root_data, graph_name, sp='val'):
     data_dict = {}
     for data_file in list_data_files:
         video_id = os.path.splitext(os.path.basename(data_file))[0]
-        
+
         with open(data_file, 'rb') as f:
             data = pickle.load(f) #nosec
 

setup.py

@@ -4,9 +4,9 @@
     long_description = f.read()
 
 setup(
-    name='graphltvu',
+    name='gravit',
     version='1.0.0',
-    description='A graph learning framework for long-term video understanding',
+    description='Graph learning framework for long-term Video undersTanding',
     long_description=long_description,
     license='Apache License 2.0',
     author='Kyle Min',

tools/evaluate.py

@@ -3,12 +3,12 @@
 import torch
 import argparse
 from torch_geometric.loader import DataLoader
-from graphltvu.utils.parser import get_cfg
-from graphltvu.utils.logger import get_logger
-from graphltvu.models import build_model
-from graphltvu.datasets import GraphDataset
-from graphltvu.utils.formatter import get_formatting_data_dict, get_formatted_preds
-from graphltvu.utils.eval_tool import get_eval_score
+from gravit.utils.parser import get_cfg
+from gravit.utils.logger import get_logger
+from gravit.models import build_model
+from gravit.datasets import GraphDataset
+from gravit.utils.formatter import get_formatting_data_dict, get_formatted_preds
+from gravit.utils.eval_tool import get_eval_score
 
 
 def evaluate(cfg):

tools/train_context_reasoning.py

@@ -3,10 +3,10 @@
 import torch
 import torch.optim as optim
 from torch_geometric.loader import DataLoader
-from graphltvu.utils.parser import get_args, get_cfg
-from graphltvu.utils.logger import get_logger
-from graphltvu.models import build_model, get_loss_func
-from graphltvu.datasets import GraphDataset
+from gravit.utils.parser import get_args, get_cfg
+from gravit.utils.logger import get_logger
+from gravit.models import build_model, get_loss_func
+from gravit.datasets import GraphDataset
 
 
 def train(cfg):