linjieli222/HERO

Issues

• Error occurs while downloading files from download_tvr.sh

  #52 opened 2 months ago by allent4n
  0

• Unable to download pretrained models & features

  #48 opened 2 years ago by Yuuraa
  2

• /scripts/download_didemo.sh

  #51 opened 2 months ago by HaungVita
  0

• The number of queries for the didemo dataset

  #50 opened 2 years ago by tfka
  0

• Unable to download pretrained models & features

  #49 opened 2 years ago by yliu-cs
  1

• How to Visualize?

  #47 opened 2 years ago by Yangaiei
  0

• Reproducing Results on TVC w/o pre-training, getting scores higher than what is reported in the paper Table 4

  #46 opened 2 years ago by Swetha5
  2

• Does this code run only on V100, which does not support 3090?

  #45 opened 3 years ago by zhangxiaoxiao-tech
  2

• Datapath

  #44 opened 3 years ago by avinashsai
  5

• Does the dimension of the input video features have to be 4352?

  #42 opened 3 years ago by onlyonewater
  2

• RuntimeError: Mismatched data types: One rank had type int64, but another rank had type uint8.

  #38 opened 3 years ago by wqf321
  2

• How long/memory cost does it take to extract all slowest features for howto100m?

  #39 opened 3 years ago by forrestsz
  2

• HPS file missing for TVQA task and eval_videoQA.py fails because of this (how to fix)?

  #43 opened 3 years ago by yrf1
  2

• Is it possible to provide the environment requirements?

  #41 opened 3 years ago by Mallory24
  1

• Reproducing VCMR results for didemo video only

  #40 opened 3 years ago by hdy007007
  1

• An error during finetuning for the TVR task

  #37 opened 3 years ago by HenryHZY
  5

• How long does it take for pre-training on TV with MLM+MNCE from scratch?

  #36 opened 3 years ago by HenryHZY
  4

• OOM in pretraining

  #30 opened 3 years ago by hgzjy25
  3

• code for how2r and how2qa

  #35 opened 3 years ago by tfka
  1

• Error while pretraining with TV data

  #33 opened 3 years ago by michaelmyc
  2

• got error while running pretrain.py

  #32 opened 3 years ago by liveseongho
  5

• I found a bug in pretrain.py

  #31 opened 3 years ago by liveseongho
  0

• some questions of model details

  #22 opened 4 years ago by JinmingZhao
  4

• The link to didemo_subtitles.db.tar file is invalid

  #20 opened 4 years ago by Unified-Robots
  4

• Pre-training based on HowTo100M dataset

  #17 opened 4 years ago by Unified-Robots
  3

• One issue about pretrain-tv-init.bin file

  #16 opened 4 years ago by Unified-Robots
  4

• The baseline code

  #15 opened 4 years ago by PipiZong
  2

• How to process the HowTo100M dataset?

  #13 opened 4 years ago by Unified-Robots
  4

• Finetuning for the TVR task

  #11 opened 4 years ago by Unified-Robots
  4

• Single-channel results on multi-channel datasets

  #2 opened 4 years ago by antoyang
  4

• question about vocab size

  #25 opened 4 years ago by taeho-kil
  2

• download HowTo100M dataset

  #24 opened 4 years ago by JKBox
  2

• Weights of HeroForVideoQA not initialized from pretrained model

  #26 opened 4 years ago by Curry-AI
  2

• Checkpoint for TVQA task

  #27 opened 4 years ago by raksharamesh14
  2

• I find a bug in svmr evaluation

  #29 opened 4 years ago by hgzjy25
  4

• TVC cannot reproduce results

  #28 opened 4 years ago by Liu0329
  2

• "lmdb.LockError: mdb_txn_begin" when using network file system.

  #21 opened 4 years ago by becxer
  6

• some questions about inplace

  #23 opened 4 years ago by y1129800378
  2

• The video features used in TVC task

  #5 opened 4 years ago by tuyunbin
  5

• Using the pretrained hero on custom datasets

  #12 opened 4 years ago by PipiZong
  0

• How many train steps are needed to get the performance of the paper when finetuning TVR dataset?

  #14 opened 4 years ago by liveseongho
  2

• Preprocessing code for VIOLIN

  #10 opened 4 years ago by Worm4047
  5

• model weights without pretraining

  #9 opened 4 years ago by kevinlee9
  1

• Can not download the video_db file

  #8 opened 4 years ago by houzhijian
  1

• Frame-level feature

  #6 opened 4 years ago by shinying
  4

• running on cpu

  #3 opened 4 years ago by Melika-Ayoughi
  2

• Feature extractor for custom dataset

  #4 opened 4 years ago by aleSuglia
  2

• when will you release the code to process the video data?

  #1 opened 4 years ago by youngfly11
  3