chevaliernoir / openasl Goto Github PK

Why does the paper not have comparison against recent SoTA benchmarks (like this and this)?
Also, why are there no results of the proposed method on the well-known Phoenix14-T benchmark?

May I ask when would the dataset and codebase be released? Currently looking for a working American Sign Language back translator. Or would it be possible for you to release the trained baseline model checkpoints in Table 2? Thanks!

In Section 3, paragraph 3, the paper states "The annotator marks the corrected beginning and end of the sentence, and provides a corrected English translation if needed as well as the corresponding gloss sequence."

BUT, when I check openasl-v1.0.tsv file, only VALID and TEST sentences have gloss annotations. Gloss entry for TRAIN set are blank. I want to ask are there some mistakes or I should get permission to have it?

In Section 3, paragraph 3, the paper states "Each annotator views the video clip and is given the original English sentence from the subtitle for reference. The annotator marks the"

1. corrected beginning and end of the sentence.
2. and provides a corrected English translation if needed

Step-2 is clear to me.

I apologize but it was not clear to me in step-1 whether the annotator

• edits the video clip timestamps to match the sentence
  OR
• edits the sentence (including/excluding words) to match what is signed in the video clip.

From Section A.7 of Appendix: When pretraining the visual encoder (I3D) model, what is the initialization of the model?
Is it initialized randomly or is it initialized with parameters trained on some video classification dataset (the way AlexNet/Inception networks used in SLT are initialized with ImageNet trained parameters).

Hi, i follow your readme, but can not download dataset, do you have another way to download dataset? e.g, google drive? baidu drive?

python prep/download.py --tsv data/openasl-v1.0.tsv
/home/arthur/PycharmProjects/OpenASL/prep/download.py:24: DtypeWarning: Columns (6) have mixed types. Specify dtype option on import or set low_memory=False.
df = pd.read_csv(tsv_fn, sep='\t')
Download 2043 raw videos into None
Traceback (most recent call last):
File "/home/arthur/PycharmProjects/OpenASL/prep/download.py", line 64, in
main()
File "/home/arthur/PycharmProjects/OpenASL/prep/download.py", line 27, in main
os.makedirs(args.dest, exist_ok=True)
File "", line 210, in makedirs
File "", line 103, in split
TypeError: expected str, bytes or os.PathLike object, not NoneType

Hi, i follow your readme, but can not download dataset, do you have another way to download dataset? e.g, google drive? baidu drive?

"[youtube] 1qMSlb1sbXw: Downloading webpage
[youtube] 1qMSlb1sbXw: Downloading MPD manifest
ERROR: Unable to extract uploader id; please report this issue on https://yt-dl.org/bug . Make sure you are using the latest version; type youtube-dl -U to update. Be sure to call youtube-dl with t
he --verbose flag and include its complete output.
4%|█████▌ | 73/2043 [04:00<1:56:01, 3.53s/it]G
:\dataset\openasl\OpenASL-main\youtube-dl 1t3Hbvflc40 -f bestvideo+bestaudio[ext=m4a]/bestvideo+bestaudio/best --merge-output-format mp4 --no-check-certificate --restrict-filenames -o 'G:/dataset/ope
nasl/OpenASL-main/data/raw-video/%(id)s.%(ext)s'
[youtube] 1t3Hbvflc40: Downloading webpage
WARNING: Unable to download webpage: HTTP Error 429: Too Many Requests
[youtube] 1t3Hbvflc40: Downloading API JSON"

Help! When I download the dataset with the following command, the error occurs:
python prep/download.py --tsv data/openasl-v1.0.tsv --dest /data1/yma1/OpenASL/raw-video/

[youtube] 00ADU7t7IWI: Downloading webpage ERROR: Unable to extract uploader id; please report this issue on https://yt-dl.org/bug . Make sure you are using the latest version; type youtube-dl -U to update. Be sure to call youtube-dl with the --verbose flag and include its complete output. 0%|▏ | 1/2043 [00:02<1:30:49, 2.67s/it] /data1/yma1/OpenASL/OpenASL/youtube-dl 03LjSYPMBK8 -f bestvideo+bestaudio[ext=m4a]/bestvideo+bestaudio/best --merge-output-format mp4 --no-check-certificate --restrict-filenames -o '/data1/yma1/OpenASL/raw-video//%(id)s.%(ext)s' [youtube] 03LjSYPMBK8: Downloading webpage ERROR: Unable to extract uploader id; please report this issue on https://yt-dl.org/bug . Make sure you are using the latest version; type youtube-dl -U to update. Be sure to call youtube-dl with the --verbose flag and include its complete output. 0%|▎ | 2/2043 [00:04<1:18:53, 2.32s/it] /data1/yma1/OpenASL/OpenASL/youtube-dl 045-gE1O_hQ -f bestvideo+bestaudio[ext=m4a]/bestvideo+bestaudio/best --merge-output-format mp4 --no-check-certificate --restrict-filenames -o '/data1/yma1/OpenASL/raw-video//%(id)s.%(ext)s' [youtube] 045-gE1O_hQ: Downloading webpage ERROR: Unable to extract uploader id; please report this issue on https://yt-dl.org/bug . Make sure you are using the latest version; type youtube-dl -U to update. Be sure to call youtube-dl with the --verbose flag and include its complete output. 0%|▍ | 3/2043 [00:07<1:18:04, 2.30s/it] /data1/yma1/OpenASL/OpenASL/youtube-dl 07BevXg9ISg -f bestvideo+bestaudio[ext=m4a]/bestvideo+bestaudio/best --merge-output-format mp4 --no-check-certificate --restrict-filenames -o '/data1/yma1/OpenASL/raw-video//%(id)s.%(ext)s' [youtube] 07BevXg9ISg: Downloading webpage ERROR: Unable to extract uploader id; please report this issue on https://yt-dl.org/bug . Make sure you are using the latest version; type youtube-dl -U to update. Be sure to call youtube-dl with the --verbose flag and include its complete output.