wmcnally / golfdb Goto Github PK
View Code? Open in Web Editor NEWGolfDB is a video database for Golf Swing Sequencing, which involves detecting 8 golf swing events in trimmed golf swing videos. This repo demos the baseline model, SwingNet.
GolfDB is a video database for Golf Swing Sequencing, which involves detecting 8 golf swing events in trimmed golf swing videos. This repo demos the baseline model, SwingNet.
RuntimeError: Error(s) in loading state_dict for MobileNetV2:
Missing key(s) in state_dict: "classifier.1.weight", "classifier.1.bias".
Unexpected key(s) in state_dict: "classifier.weight", "classifier.bias".
Hello,
when i am running "python train.py ",, i have this message in below: " **training, momentum, eps, torch.backends.cudnn.enabled
**RuntimeError: CUDA out of memory. Tried to allocate 1.07 GiB (GPU 0; 4.00 GiB total capacity; 2.57 GiB already allocated; 84.45 MiB free; 2.59 GiB reserved in total by PyTorch)****"
Any idea what might cause this ?
Hi,
I want to use this model to return the location of the swing events for an individual video or even better, an image of each of the events. How should I modify eval.py to do this?
I also don't have access to a computer with a GPU at the moment because of Covid-19, so is it possible to do this without CUDA?
Kind regards,
Liam
Hi,
Are the original full resolution videos available for download (instead of the 160x160 downscaled)?
Thx!
Wouter
I am reading your paper entitled "GolfDB: A Video Database for Golf Swing Sequencing" saying that there will be a curated list of high quality golf sequencing videos for benchmarks. So when will the videos be available?
/content/drive/MyDrive/golfdb_master# python test_video.py -p content/drive/MyDrive/golfdb_master/test_video.mp4
Preparing video: content/drive/MyDrive/golfdb_master/test_video.mp4
Using device: cuda
Loaded model weights
Testing...
Traceback (most recent call last):
File "test_video.py", line 95, in
for sample in dl:
File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py", line 517, in next
data = self._next_data()
File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py", line 557, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/fetch.py", line 44, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "test_video.py", line 35, in getitem
ratio = self.input_size / max(frame_size)
ZeroDivisionError: float division by zero
Hello,
Everything works fine,, but when i aml trying to run "python test_video.p -p test_video.mp4 ",, i have this message in below: zeroDevisionError..
Did you happen to solve the problem for yourself? What did it end up being?
How can I train my model on custom videos that I have with me? Can you help me out?
I'm interesting on your paper:GolfDB:AVideoDatabaseforGolfSwingSequencing.
So I want to request your SwingNet code for research purpose.
Thanks a lot.
Hi Will,
Thank you so much for making your dataset and code publicly available. May I ask you for the unprocessed video clips? The reason I am asking this is that I am trying to learn the texture and normal map for golf players from the videos. My project will be based on your high-resolution video set. Thank you again for your help.
Zach
Hello,
I was hoping to re-train golfdb on a larger video size dataset and the I can't find the URLS for the youtube videos in the data set folder. Are the urls available somewhere?
Hello!
I want to know what is the seq_lenghth in train.py and eval.py?
thanks a lot!
Is there any problem in copyright even if you download or use YouTube Golf Swing data, GolfDB?
Hello!
I want to ask about average time for event detection.
Videos times are between 7 seconds and 10 seconds.. every how much time this event detection happen?, I mean the average of time between each events detection.
thanks in advance
I am in need of the videos in higher resolution and it has been mentioned that we can use different input configuration by downloading the youtube videos via the URLs provided in the data set and pre-processing them ourselves. But I could not find the URLs anywhere.
I want to load the videos by myself, but I can't load the pickle file with the video addresses. The error is given as: AttributeError: Can't get attribute 'new_block' on <module 'pandas.core.internals.blocks'
.
Hope there is a solution to this problem.
Thank you in advance
Hello,
There is a problem with the installation of pytorch, so I want to execute pip install pytorch==1.3 alone, but the system prompts no Found that pytorch==1.3, and then I checked on the pytorch official website, I did not find pytorch 1.3 version, so I want to confirm whether there is a version error in the environment you provide?
I'm interested in your code, so I'm trying to check the pretrained model, but emurate(data_loader) doesn't enter. Do you know how to fix it?
I did as you said and changed the train of the dataset to True.
Hey i am not able to get the coordinates of golf face even from test_video.py. Can you please tell me how do i get coordinates and not just frames where golf face is detected?
Originally posted by @onkar-99 in #5 (comment)
I want to play eval.py with my golf video.
But I can't figure out how to make a mat file like golfdb so it's impossible to test it with my golf video.
I tried to use preprocess_videos.py but it failed.
Can you teach me one example of preprocess_videos.py and mat files?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.