Fictionarry/TalkingGaussian

Issues

• Training with longer videos

  #49 opened 3 months ago by Turlan
  2

• csv

  #48 opened 3 months ago by Baixue1234
  1

• 推理时输入全是0的音频，嘴依然是微张的，这个有没有办法解决呢？

  #21 opened 6 months ago by anliyuan
  2

• ModuleNotFoundError: No module named 'diff_gaussian_rasterization'

  #47 opened 3 months ago by xiaomaofei
  2

• How can i run it in realtime ?

  #45 opened 3 months ago by Deltaidiots
  1

• cannot clone repo...

  #46 opened 3 months ago by Pixie8888
  1

• Is this better than synctalk?

  #44 opened 4 months ago by linhcentrio
  2

• 牙齿周围伪影问题

  #43 opened 4 months ago by wwqy
  3

• [windows 10, cuda 11.8] TypeError: grid_encode_forward() Incompatible Function Arguments:, FileNotFoundError: chkpnt_face_latest.pth, ZeroDivisionError: division by zero

  #42 opened 4 months ago by linhcentrio
  2

• 训练走到train_face 报错

  #31 opened 4 months ago by 8600862
  3

• ZeroDivisionError during training

  #41 opened 4 months ago by MahlerMozart
  1

• 在新的推理上表现较差

  #40 opened 4 months ago by Jerry-sjw
  1

• 有没交流群，可以给我学习一下，交流的一下的地方？

  #39 opened 4 months ago by NuoLanC
  0

• Training Results Not Matching Demo Quality – Possible Overclaim?

  #38 opened 4 months ago by caliber1313
  4

• python data_utils/process.py data/may/may.mp4怎么要运行那么久

  #37 opened 4 months ago by qingtianwu
  0

• 是否能开源躯干训练代码

  #36 opened 4 months ago by xiaoqidaov2
  0

• 请问如果想在渲染时更换背景，该怎么操作呢？

  #35 opened 4 months ago by Coooderr
  2

• 请问大佬推理目标语音的结果在哪里？

  #29 opened 5 months ago by zhanghongyong123456
  3

• train_face时 warm_up结束后loss变成nan，训练结束后测试，人非常抖

  #17 opened 6 months ago by anliyuan
  3

• NaN or Inf found in input tensor

  #34 opened 5 months ago by xiaoqidaov2
  1

• 噪点问题解决

  #33 opened 5 months ago by THE-120
  0

• 修改tracking后生成的视频出现了很多问题

  #30 opened 5 months ago by xiaoqidaov2
  2

• 请问嘴型对齐的效果还能提升吗？

  #32 opened 5 months ago by charryshi
  3

• 牙齿区域检测不准问题

  #4 opened 6 months ago by yulj21
  15

• synthesize_fuse.py跑出来的视频out.mp4没有头，有没有一样情况的？

  #27 opened 5 months ago by lxh0319
  1

• train_face.py训练后期报错：Given input size: (192x2x2). Calculated output size: (192x0x0). Output size is too small

  #26 opened 5 months ago by SherlockSunset
  8

• 如果用hubert的话，有办法能流式调吗？

  #28 opened 5 months ago by anliyuan
  0

• bash scripts/train_xx.sh，训练时报错，是怎么回事呢

  #24 opened 5 months ago by lxh0319
  2

• FileNotFoundError and ZeroDivisionError during training

  #25 opened 5 months ago by YanMinMacMaster
  4

• ER-Nerf中的checkpoints/<ngp_ep0043.pth>在TalkingGaussian中没有找到，是否取消了checkpoints？

  #23 opened 5 months ago by YPwan
  4

• RuntimeError: CUDA error: CUBLAS_STATUS_INVALID_VALUE when calling `cublasSgemm( handle, opa, opb, m, n, k, &alpha, a, lda, b, ldb, &beta, c, ldc)`

  #14 opened 6 months ago by einsqing
  7

• How to calculate the AUE-(L/U)

  #22 opened 6 months ago by Net-Maker
  2

• Cannot finish pre-processing video

  #16 opened 6 months ago by rawrshah
  1

• 视频预处理的第8步，提示说粗糙栅格化阶段的 bin 大小太小，导致了溢出

  #19 opened 6 months ago by Ivyforever01
  1

• 中文口型对不上

  #20 opened 6 months ago by zhouzhenneng
  7

• 推理后只有脸，其他部分全是黑的，没有背景

  #9 opened 6 months ago by anliyuan
  3

• Will it be better than ER-NeRF in terms of inferential efficiency ?

  #18 opened 6 months ago by 2033723741
  2

• train_mouth.py训练报错

  #15 opened 6 months ago by yulj21
  3

• 请问中文语音的唇形同步，是否跟ER-NeRF一样，使用hubert特征会更好？

  #12 opened 6 months ago by husthzy
  1

• What's the PyMCube version?

  #13 opened 6 months ago by Net-Maker
  2

• simple-knn not found

  #11 opened 6 months ago by Saksham209
  1

• 数据预处理错误 python3 data_utils/process.py /app/TalkingGaussian/data/May/May.mp4 --task 8

  #8 opened 6 months ago by Tiandishihua
  1

• inference time

  #10 opened 6 months ago by Tiandishihua
  1

• not enough values to unpack (expected 4, got 3)

  #7 opened 6 months ago by caitongbo
  3

• wav2vec still working?

  #6 opened 6 months ago by tngt
  1

• hubert features code

  #5 opened 6 months ago by trinhtuanvubk
  1

• Code Release Estimate

  #3 opened 7 months ago by nazmicancalik
  3

• Training code / preprocessing

  #2 opened 6 months ago by johndpope
  2

• Code Open Source Plan Inquiry

  #1 opened 9 months ago by Kedreamix
  2