FangSen9000 commited on
Commit ·
1bbfe4e
1
Parent(s): 0c05e55
Add a visual gallery demonstration
Browse files- README.md +43 -1
- generated_pose_video/HELLO_WORLD_27184_63831_20250803_235543.mp4 +3 -0
- generated_pose_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750.mp4 +3 -0
- generated_pose_video/NOT_FIND_YET_38868_21853_38946_20250803_235630.mp4 +3 -0
- generated_pose_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927.mp4 +3 -0
- generated_sign_video/HELLO_WORLD_27184_63831_20250803_235543.mp4 +3 -0
- generated_sign_video/HELLO_WORLD_27184_63831_20250803_235543_combined.mp4 +3 -0
- generated_sign_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750.mp4 +3 -0
- generated_sign_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750_combined.mp4 +3 -0
- generated_sign_video/NOT_FIND_YET_38868_21853_38946_20250803_235630.mp4 +3 -0
- generated_sign_video/NOT_FIND_YET_38868_21853_38946_20250803_235630_combined.mp4 +3 -0
- generated_sign_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927.mp4 +3 -0
- generated_sign_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927_combined.mp4 +3 -0
README.md
CHANGED
|
@@ -4,7 +4,7 @@ The entire project is divided into two parts: `text_or_gloss2pose` and `pose2vid
|
|
| 4 |
|
| 5 |
- In the text2pose section (`Expected 4 seconds / one video`), I first used the DWpose tool to extract the npz format pose information for each word video frame by frame, and then merged each frame's npz into a large npz file. That is to say, we know the consecutive poses corresponding to each word. These consecutive poses are stored in the merged npz file. (We can also use deep learning models to generate these poses, but for commercial-level accuracy, this original data Dict method will be much more accurate.)
|
| 6 |
|
| 7 |
-
- In the pose2video section (`Expected
|
| 8 |
|
| 9 |
# For Users
|
| 10 |
|
|
@@ -41,6 +41,48 @@ CUDA_VISIBLE_DEVICES=0 python ControlNeXt-SVD-v2/run_controlnext_optimized.py \
|
|
| 41 |
--decode_chunk_size 2
|
| 42 |
```
|
| 43 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
|
| 45 |
# For developers
|
| 46 |
|
|
|
|
| 4 |
|
| 5 |
- In the text2pose section (`Expected 4 seconds / one video`), I first used the DWpose tool to extract the npz format pose information for each word video frame by frame, and then merged each frame's npz into a large npz file. That is to say, we know the consecutive poses corresponding to each word. These consecutive poses are stored in the merged npz file. (We can also use deep learning models to generate these poses, but for commercial-level accuracy, this original data Dict method will be much more accurate.)
|
| 6 |
|
| 7 |
+
- In the pose2video section (`Expected 6 minutes / one video`), we render the visualized poses of the skeletons generated by each word. I used the ControlNext-SVD model for this.
|
| 8 |
|
| 9 |
# For Users
|
| 10 |
|
|
|
|
| 41 |
--decode_chunk_size 2
|
| 42 |
```
|
| 43 |
|
| 44 |
+
## Demo Gallery
|
| 45 |
+
|
| 46 |
+
<details>
|
| 47 |
+
<summary>🎬 Click to view generated examples</summary>
|
| 48 |
+
|
| 49 |
+
### Example 1: "good morning afternoon"
|
| 50 |
+
- **Input text**: `good morning afternoon`
|
| 51 |
+
- **Pose video**: `generated_pose_video/GOOD_MORNING_AFTERNOON_25073_36863_01435_20250803_233215.mp4`
|
| 52 |
+
- **Final video**: `generated_sign_video/GOOD_MORNING_AFTERNOON_25073_36863_01435_20250803_233215.mp4`
|
| 53 |
+
- **Combined view**: `generated_sign_video/GOOD_MORNING_AFTERNOON_25073_36863_01435_20250803_233215_combined.mp4`
|
| 54 |
+
|
| 55 |
+
### Example 2: "hello world"
|
| 56 |
+
- **Input text**: `hello world`
|
| 57 |
+
- **Pose video**: `generated_pose_video/HELLO_WORLD_27184_63831_20250803_235543.mp4`
|
| 58 |
+
- **Final video**: `generated_sign_video/HELLO_WORLD_27184_63831_20250803_235543.mp4`
|
| 59 |
+
- **Combined view**: `generated_sign_video/HELLO_WORLD_27184_63831_20250803_235543_combined.mp4`
|
| 60 |
+
|
| 61 |
+
### Example 3: "I will attend this dinner party"
|
| 62 |
+
- **Input text**: `I will attend this dinner party`
|
| 63 |
+
- **Pose video**: `generated_pose_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750.mp4`
|
| 64 |
+
- **Final video**: `generated_sign_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750.mp4`
|
| 65 |
+
- **Combined view**: `generated_sign_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750_combined.mp4`
|
| 66 |
+
|
| 67 |
+
### Example 4: "not find yet"
|
| 68 |
+
- **Input text**: `not find yet`
|
| 69 |
+
- **Pose video**: `generated_pose_video/NOT_FIND_YET_38868_21853_38946_20250803_235630.mp4`
|
| 70 |
+
- **Final video**: `generated_sign_video/NOT_FIND_YET_38868_21853_38946_20250803_235630.mp4`
|
| 71 |
+
- **Combined view**: `generated_sign_video/NOT_FIND_YET_38868_21853_38946_20250803_235630_combined.mp4`
|
| 72 |
+
|
| 73 |
+
### Example 5: "where is my cat"
|
| 74 |
+
- **Input text**: `where is my cat`
|
| 75 |
+
- **Pose video**: `generated_pose_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927.mp4`
|
| 76 |
+
- **Final video**: `generated_sign_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927.mp4`
|
| 77 |
+
- **Combined view**: `generated_sign_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927_combined.mp4`
|
| 78 |
+
|
| 79 |
+
### Note on video types:
|
| 80 |
+
- **Pose video**: Shows the skeleton/pose generated from input text/gloss
|
| 81 |
+
- **Final video**: The rendered sign language video with a virtual signer
|
| 82 |
+
- **Combined view**: Side-by-side comparison of pose and final video
|
| 83 |
+
|
| 84 |
+
</details>
|
| 85 |
+
|
| 86 |
|
| 87 |
# For developers
|
| 88 |
|
generated_pose_video/HELLO_WORLD_27184_63831_20250803_235543.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:653eaa3cb9d77210c52f44041e3c691dcf790ac91ab3c2e021d5b105ca048bc4
|
| 3 |
+
size 2054014
|
generated_pose_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4b5847ee1c412ed6908732736db1960fbbf93fbc13e1cf223a66bfac235226d
|
| 3 |
+
size 4383086
|
generated_pose_video/NOT_FIND_YET_38868_21853_38946_20250803_235630.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd1590173ed3144ff3deb53615fd270ad5a7509053adb6b6a284a8acf79ea689
|
| 3 |
+
size 2410241
|
generated_pose_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85c5f6d2002e4bc2f495c05a5bd5f47442b3e4fd02cec76e28e05c4a8b1e09ed
|
| 3 |
+
size 3186550
|
generated_sign_video/HELLO_WORLD_27184_63831_20250803_235543.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad825761d1b15f4b22ab36c3fdff617b2a65141a9f10fc3d8603811ce52a4868
|
| 3 |
+
size 1333936
|
generated_sign_video/HELLO_WORLD_27184_63831_20250803_235543_combined.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a44757adab0f12e339fe2f11c65fe2a018e6573193619bfec944ee121a977f3
|
| 3 |
+
size 1773622
|
generated_sign_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3868939e530760666e17ce8d57f899a08e2c52b045bf8c078d222f8864599af0
|
| 3 |
+
size 2387083
|
generated_sign_video/I_WILL_ATTEND_THIS_DINNER_PARTY_28794_63364_04007_66628_16327_41325_20250803_235750_combined.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95f7ad4716cc667a0b66fddd45605c6b27f2e5bacc0e9c8380621d1869fc4abb
|
| 3 |
+
size 3168754
|
generated_sign_video/NOT_FIND_YET_38868_21853_38946_20250803_235630.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ffe57314a6c0dac56f8e8094533e885ba392e080b136ac832bc8f0117125f458
|
| 3 |
+
size 1601408
|
generated_sign_video/NOT_FIND_YET_38868_21853_38946_20250803_235630_combined.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:814269dc74751c20f1b68d8c8d17ef9336b7ffc3c0c350a1e3986c690be07f25
|
| 3 |
+
size 2107461
|
generated_sign_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:20ae3d6c65f86393cab08e32f9380f37af5a8a26a46129df6569787ae92dcb26
|
| 3 |
+
size 2015616
|
generated_sign_video/WHERE_IS_MY_CAT_63084_58616_37469_65313_20250803_235927_combined.mp4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:071749297784b453bda19fd6bb5cd5c679056a2e5dbcedcd24997f0c525d64de
|
| 3 |
+
size 2693309
|