camenduru/hand-refiner 🔢📝🖼️ → 🖼️

▶️ 1.5K runs 📅 Feb 2024 ⚙️ Cog 0.9.4 🔗 GitHub 📄 Paper ⚖️ License
hand-refinement image-enhancement image-processing portrait-enhancement

About

Hand Refiner 512x512

Example Output

Prompt:

"a person facing the camera, making a hand gesture, indoor"

Output

Example output

Performance Metrics

21.15s Prediction Time
224.59s Total Time
All Input Parameters
{
  "seed": 34343,
  "prompt": "a person facing the camera, making a hand gesture, indoor",
  "strength": 0.6,
  "input_image": "https://replicate.delivery/pbxt/KMZkbywX0JCSzdTmeUl3VsQucok0tFXrLeTXrnASvn4Tkp65/1.jpg"
}
Input Parameters
seed Type: integerDefault: 1
prompt Type: stringDefault: a person facing the camera, making a hand gesture, indoor
strength Type: numberDefault: 0.55Range: 0 - 1
Strength
input_image (required) Type: string
Input Image
Output Schema

Output

Type: stringFormat: uri

Example Execution Logs
set os.environ[OMP_NUM_THREADS] to 4
/content/HandRefiner/MeshGraphormer/manopth/manopth/manolayer.py:72: UserWarning: The given NumPy array is not writable, and PyTorch does not support non-writable tensors. This means writing to this tensor will result in undefined behavior. You may want to copy the array to protect its data or make it writable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at ../torch/csrc/utils/tensor_numpy.cpp:206.)
torch.Tensor(smpl_data['betas'].r).unsqueeze(0))
/content/HandRefiner/MeshGraphormer/src/modeling/_mano.py:94: UserWarning: torch.sparse.SparseTensor(indices, values, shape, *, device=) is deprecated.  Please use torch.sparse_coo_tensor(indices, values, shape, dtype=, device=). (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:605.)
ptU.append(torch.sparse.FloatTensor(i, v, u.shape))
/usr/local/lib/python3.10/site-packages/scipy/sparse/_index.py:102: SparseEfficiencyWarning: Changing the sparsity structure of a csr_matrix is expensive. lil_matrix is more efficient.
self._set_intXint(row, col, x.flat[0])
=> loading pretrained model MeshGraphormer/models/hrnet/hrnetv2_w64_imagenet_pretrained.pth
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1707546322.029188     119 gl_context_egl.cc:85] Successfully initialized EGL. Major : 1 Minor: 5
I0000 00:00:1707546322.051412     331 gl_context.cc:344] GL version: 3.2 (OpenGL ES 3.2 NVIDIA 525.147.05), renderer: NVIDIA A40/PCIe/SSE2
INFO: Created TensorFlow Lite XNNPACK delegate for CPU.
Seed set to 34343
Seed set to 34343
Data shape for DDIM sampling is (1, 4, 64, 64), eta 0.0
Running DDIM Sampling with 50 timesteps
DDIM Sampler:   0%|          | 0/50 [00:00<?, ?it/s]
DDIM Sampler:   2%|▏         | 1/50 [00:00<00:12,  3.81it/s]
DDIM Sampler:   4%|▍         | 2/50 [00:00<00:10,  4.69it/s]
DDIM Sampler:   6%|▌         | 3/50 [00:00<00:09,  5.06it/s]
DDIM Sampler:   8%|▊         | 4/50 [00:00<00:08,  5.26it/s]
DDIM Sampler:  10%|█         | 5/50 [00:00<00:08,  5.37it/s]
DDIM Sampler:  12%|█▏        | 6/50 [00:01<00:08,  5.44it/s]
DDIM Sampler:  14%|█▍        | 7/50 [00:01<00:07,  5.49it/s]
DDIM Sampler:  16%|█▌        | 8/50 [00:01<00:07,  5.52it/s]
DDIM Sampler:  18%|█▊        | 9/50 [00:01<00:07,  5.54it/s]
DDIM Sampler:  20%|██        | 10/50 [00:01<00:07,  5.54it/s]
DDIM Sampler:  22%|██▏       | 11/50 [00:02<00:07,  5.55it/s]
DDIM Sampler:  24%|██▍       | 12/50 [00:02<00:06,  5.56it/s]
DDIM Sampler:  26%|██▌       | 13/50 [00:02<00:06,  5.56it/s]
DDIM Sampler:  28%|██▊       | 14/50 [00:02<00:06,  5.57it/s]
DDIM Sampler:  30%|███       | 15/50 [00:02<00:06,  5.57it/s]
DDIM Sampler:  32%|███▏      | 16/50 [00:02<00:06,  5.57it/s]
DDIM Sampler:  34%|███▍      | 17/50 [00:03<00:05,  5.57it/s]
DDIM Sampler:  36%|███▌      | 18/50 [00:03<00:05,  5.57it/s]
DDIM Sampler:  38%|███▊      | 19/50 [00:03<00:05,  5.56it/s]
DDIM Sampler:  40%|████      | 20/50 [00:03<00:05,  5.57it/s]
DDIM Sampler:  42%|████▏     | 21/50 [00:03<00:05,  5.56it/s]
DDIM Sampler:  44%|████▍     | 22/50 [00:04<00:05,  5.57it/s]
DDIM Sampler:  46%|████▌     | 23/50 [00:04<00:04,  5.57it/s]
DDIM Sampler:  48%|████▊     | 24/50 [00:04<00:04,  5.57it/s]
DDIM Sampler:  50%|█████     | 25/50 [00:04<00:04,  5.56it/s]
DDIM Sampler:  52%|█████▏    | 26/50 [00:04<00:04,  5.57it/s]
DDIM Sampler:  54%|█████▍    | 27/50 [00:04<00:04,  5.57it/s]
DDIM Sampler:  56%|█████▌    | 28/50 [00:05<00:03,  5.57it/s]
DDIM Sampler:  58%|█████▊    | 29/50 [00:05<00:03,  5.57it/s]
DDIM Sampler:  60%|██████    | 30/50 [00:05<00:03,  5.57it/s]
DDIM Sampler:  62%|██████▏   | 31/50 [00:05<00:03,  5.57it/s]
DDIM Sampler:  64%|██████▍   | 32/50 [00:05<00:03,  5.57it/s]
DDIM Sampler:  66%|██████▌   | 33/50 [00:06<00:03,  5.57it/s]
DDIM Sampler:  68%|██████▊   | 34/50 [00:06<00:02,  5.57it/s]
DDIM Sampler:  70%|███████   | 35/50 [00:06<00:02,  5.57it/s]
DDIM Sampler:  72%|███████▏  | 36/50 [00:06<00:02,  5.57it/s]
DDIM Sampler:  74%|███████▍  | 37/50 [00:06<00:02,  5.57it/s]
DDIM Sampler:  76%|███████▌  | 38/50 [00:06<00:02,  5.56it/s]
DDIM Sampler:  78%|███████▊  | 39/50 [00:07<00:01,  5.56it/s]
DDIM Sampler:  80%|████████  | 40/50 [00:07<00:01,  5.56it/s]
DDIM Sampler:  82%|████████▏ | 41/50 [00:07<00:01,  5.56it/s]
DDIM Sampler:  84%|████████▍ | 42/50 [00:07<00:01,  5.56it/s]
DDIM Sampler:  86%|████████▌ | 43/50 [00:07<00:01,  5.56it/s]
DDIM Sampler:  88%|████████▊ | 44/50 [00:07<00:01,  5.57it/s]
DDIM Sampler:  90%|█████████ | 45/50 [00:08<00:00,  5.57it/s]
DDIM Sampler:  92%|█████████▏| 46/50 [00:08<00:00,  5.57it/s]
DDIM Sampler:  94%|█████████▍| 47/50 [00:08<00:00,  5.57it/s]
DDIM Sampler:  96%|█████████▌| 48/50 [00:08<00:00,  5.57it/s]
DDIM Sampler:  98%|█████████▊| 49/50 [00:08<00:00,  5.56it/s]
DDIM Sampler: 100%|██████████| 50/50 [00:09<00:00,  5.56it/s]
DDIM Sampler: 100%|██████████| 50/50 [00:09<00:00,  5.52it/s]
Version Details
Version ID
a1314c7fb83fec0d6ef991dfd97ae671198ea3c793dc42a01f59d92e6ce472d7
Version Created
February 9, 2024
Run on Replicate →