Updating dependency

.gitignore

@@ -0,0 +1 @@
+.idea/

Base-RCNN-FPN.yaml

@@ -0,0 +1,42 @@
+MODEL:
+  META_ARCHITECTURE: "GeneralizedRCNN"
+  BACKBONE:
+    NAME: "build_resnet_fpn_backbone"
+  RESNETS:
+    OUT_FEATURES: ["res2", "res3", "res4", "res5"]
+  FPN:
+    IN_FEATURES: ["res2", "res3", "res4", "res5"]
+  ANCHOR_GENERATOR:
+    SIZES: [[32], [64], [128], [256], [512]]  # One size for each in feature map
+    ASPECT_RATIOS: [[0.5, 1.0, 2.0]]  # Three aspect ratios (same for all in feature maps)
+  RPN:
+    IN_FEATURES: ["p2", "p3", "p4", "p5", "p6"]
+    PRE_NMS_TOPK_TRAIN: 2000  # Per FPN level
+    PRE_NMS_TOPK_TEST: 1000  # Per FPN level
+    # Detectron1 uses 2000 proposals per-batch,
+    # (See "modeling/rpn/rpn_outputs.py" for details of this legacy issue)
+    # which is approximately 1000 proposals per-image since the default batch size for FPN is 2.
+    POST_NMS_TOPK_TRAIN: 1000
+    POST_NMS_TOPK_TEST: 1000
+  ROI_HEADS:
+    NAME: "StandardROIHeads"
+    IN_FEATURES: ["p2", "p3", "p4", "p5"]
+  ROI_BOX_HEAD:
+    NAME: "FastRCNNConvFCHead"
+    NUM_FC: 2
+    POOLER_RESOLUTION: 7
+  ROI_MASK_HEAD:
+    NAME: "MaskRCNNConvUpsampleHead"
+    NUM_CONV: 4
+    POOLER_RESOLUTION: 14
+DATASETS:
+  TRAIN: ("coco_2017_train",)
+  TEST: ("coco_2017_val",)
+SOLVER:
+  IMS_PER_BATCH: 16
+  BASE_LR: 0.02
+  STEPS: (60000, 80000)
+  MAX_ITER: 90000
+INPUT:
+  MIN_SIZE_TRAIN: (640, 672, 704, 736, 768, 800)
+VERSION: 2

app.py

@@ -1,22 +1,28 @@
+import os 
+os.system('pip install git+https://github.com/facebookresearch/detectron2.git')
+
 import gradio as gr
 import logging
 from detectron2.engine import DefaultPredictor
 import cv2
 from detectron2.config import get_cfg
-from src.utils.visualizer import add_bboxes
+from utils import add_bboxes
+
+# print(torch.__version__, torch.cuda.is_available())
+# assert torch.__version__.startswith("1.9") 
 
 config_file="config.yaml"
 cfg = get_cfg()
 cfg.merge_from_file(config_file)
 cfg.MODEL.DEVICE="cpu"
 cfg.MODEL.ROI_HEADS.SCORE_THRESH_TEST = 0.5
-cfg.MODEL.WEIGHTS = "checkpoints_model_final_imagenet_40k_synthetic.pth.pth"
+cfg.MODEL.WEIGHTS = "checkpoints_model_final_imagenet_40k_synthetic.pth"
 
 def predict(
-    config_file, checkpoint_file, img_path
+    img
 ):
     predictor = DefaultPredictor(cfg)
-    im = cv2.imread(img_path)
+    im = cv2.imread(img.name)
     output = predictor(im)
     img = add_bboxes(im, output['instances'].pred_boxes, scores=output['instances'].scores)
     return img

config.yaml

@@ -0,0 +1,22 @@
+_BASE_: "./Base-RCNN-FPN.yaml"
+MODEL:
+  WEIGHTS: "detectron2://ImageNetPretrained/MSRA/R-50.pkl"
+  MASK_ON: False
+  BACKBONE:
+    FREEZE_AT: 0
+  RESNETS:
+    DEPTH: 50
+  ROI_HEADS:
+    NUM_CLASSES: 1
+SOLVER:
+  STEPS: (210000, 250000)
+  MAX_ITER: 270000
+  IMS_PER_BATCH: 1
+  BASE_LR: 0.00025
+  GAMMA: 0.1
+  REFERENCE_WORLD_SIZE: 1
+
+DATALOADER:
+  NUM_WORKERS: 1
+
+

requirements.txt

@@ -1,4 +1,25 @@
+pyyaml==5.1
 torch==1.9.0
 torchvision==0.10.0
-git+https://github.com/facebookresearch/detectron2.git
-opencv-python
+
+docutils==0.16
+# https://github.com/sphinx-doc/sphinx/commit/7acd3ada3f38076af7b2b5c9f3b60bb9c2587a3d
+sphinx==3.2.0
+recommonmark==0.6.0
+sphinx_rtd_theme
+# Dependencies here are only those required by import
+termcolor
+numpy
+tqdm
+matplotlib
+termcolor
+yacs
+tabulate
+cloudpickle
+Pillow
+future
+git+https://github.com/facebookresearch/fvcore
+omegaconf>=2.1.0.dev24
+hydra-core>=1.1.0.dev5
+
+opencv-python-headless