omnimatte
omnimatte copied to clipboard
erikalu
I could almostly reproduce the results when using the provided H file, but failed to get the right results when using the H matrix calculated by myself using the OpenCV...
When I start train script for " python train.py --name tennis --dataroot ./datasets/tennis --gpu_ids 0" I get an error for connect server  and block with this step 
If so, can somebody please share a step-by-step guide how to install it / make it work ? Thanks ahead!
Hi Erika, thanks for sharing this nice work. I'm trying to understand the alpha regularization term and I've found what I believe to be either a bug or a discrepancy...
Hello, trying to pip install requirements but getting this error: ```!cd omnimatte && ls && pip3 install -r requirements.txt``` ``` Collecting torch==1.8.0 Using cached torch-1.8.0-cp37-cp37m-manylinux1_x86_64.whl (735.5 MB) Requirement already satisfied:...
According to the 2th step of processing our own video, we should resize the video to 256x448 and save the frames in my_video/rgb. But at the 6th step, register the...
Would you please give detail guidline or example of how to get the forward/backword .flo files from RAFT?
  ``` im = cv2.imread("./1.jpg") cv2_imshow(im) cfg = get_cfg() cfg.merge_from_file(model_zoo.get_config_file("COCO-InstanceSegmentation/mask_rcnn_R_50_FPN_3x.yaml")) cfg.MODEL.ROI_HEADS.SCORE_THRESH_TEST = 0.5 cfg.MODEL.WEIGHTS = model_zoo.get_checkpoint_url("COCO-InstanceSegmentation/mask_rcnn_R_50_FPN_3x.yaml") predictor = DefaultPredictor(cfg) outputs = predictor(im) v = Visualizer(im[:, :, ::-1], MetadataCatalog.get(cfg.DATASETS.TRAIN[0]), scale=1.2)...
It would be great to see a google colab / jupyter notebook of omnimatte
Could you please upload the trained model for the tennis example (and possibly other examples) with the data so we don't have to retrain the models just to test it...
