SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2302.05496 · CVPR · 2023

MaskSketch: Unpaired Structure-guided Masked Image Generation

Kihyuk Sohn, Kate Saenko, Dina Bashkirova, Irfan Essa, José Lezama

arXiv · PDF · Open in the Atlas

Code that ran

12 of 15 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 9 have no contract verdict; 2 failures of our instrument; 3 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2302.05496.json · badge: /api/ran/2302.05496.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
google-research/masksketch canonical (classifier label)
classifier basis not recorded on the link
12 ran · 3 no recorded run
FunctionStatusWhere it lives
apply_label_smoothing Ran
ran, violated the contract we drafted
google-research/masksketch/masksketch/libml/losses.py
code served (permissive licence) · get_code("1b2e0a7656cc76a7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ArrayImpl, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import jax.numpy as jnp
import numpy as np

# Create a small one-hot encoded targets array
# Shape: [batch_size=2, num_classes=4]
batch_size = 2
num_classes = 4
one_hot_targets = jnp.array([
    [1.0, 0.0, 0.0, 0.0],
    [0.0, 1.0, 0.0, 0.0]
], dtype=jnp.float32)

# Label smoothing parameter (scalar in [0, 1])
label_smoothing = 0.1

ARGS = {
    "one_hot_targets": one_hot_targets,
    "label_smoothing": label_smoothing
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_hyper Ran
ran; our contract draft was wrong, not the code
google-research/masksketch/masksketch/configs/base_config.py
code served (permissive licence) · get_code("ffc38a27fd9393ba")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

class MockHyperparameter:
    def __init__(self):
        self.sweeps = {}
        self.products = []
    
    def sweep(self, name, values):
        """Create a sweep over values"""
        self.sweeps[name] = values
        return {name: values}
    
    def product(self, sweeps, name=None):
        """Create a product of sweeps"""
        result = []
        for sweep in sweeps:
            if isinstance(sweep, dict):
                for key, values in sweep.items():
                    for value in values:
                        result.append({key: value})
        return result

h = MockHyperparameter()

ARGS = {
    'h': h
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_hyper Ran
ran; our contract draft was wrong, not the code
google-research/masksketch/masksketch/configs/maskgit_class_cond_config.py
code served (permissive licence) · get_code("7dfd4d71b4b83843")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn


class HyperparameterConfig:
    """Stand-in class for hyperparameter configuration object."""
    
    def __init__(self):
        self.params = {}
    
    def sweep(self, key, values):
        """Mock sweep method that stores sweep configuration."""
        return {'key': key, 'values': values}
    
    def product(self, sweeps, name=None):
        """Mock product method that combines sweeps."""
        return {
            'type': 'product',
            'sweeps': sweeps,
            'name': name
        }


h = HyperparameterConfig()

ARGS = {
    'h': h
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_hyper Ran
ran on a synthesized input
google-research/masksketch/masksketch/configs/vqgan_config.py
code served (permissive licence) · get_code("cd108d44d16c6e3c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class MockHyperObject:
    """Mock object that has a product method"""
    def product(self, keys, name=None):
        """Returns a dict with the given name as key"""
        return {name: "config_value"}

h = MockHyperObject()

ARGS = {
    'h': h
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_kqv Ran
ran on a synthesized input
google-research/masksketch/masksketch/sketch_conditional_inference.py
code served (permissive licence) · get_code("dfa11013d8646365")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import jax.numpy as jnp
import numpy as np

# Build the intermediates dict structure that get_kqv expects
intermediates = {
    'TransformerLayer_0': {
        'Attention_0': {
            'self_attention': {
                'key': {
                    '__call__': [np.arange(4, dtype=np.float32)]
                },
                'value': {
                    '__call__': [np.arange(4, 8, dtype=np.float32)]
                },
                'query': {
                    '__call__': [np.arange(8, 12, dtype=np.float32)]
                }
            }
        }
    },
    'TransformerLayer_1': {
        'Attention_0': {
            'self_attention': {
                'key': {
                    '__call__': [np.arange(12, 16, dtype=np.float32)]
                },
                'value': {
                    '__call__': [np.arange(16, 20, dtype=np.float32)]
                },
                'query': {
                    '__call__': [np.arange(20, 24, dtype=np.float32)]
                }
            }
        }
    },
    'TransformerLayer_2': {
        'Attention_0': {
            'self_attention': {
                'key': {
                    '__call__': [np.arange(24, 28, dtype=np.float32)]
                },
                'value': {
                    '__call__': [np.arange(28, 32, dtype=np.float32)]
                },
                'query': {
                    '__call__': [np.arange(32, 36, dtype=np.float32)]
                }
            }
        }
    }
}

num_layers = 3

ARGS = {
    'intermediates': intermediates,
    'num_layers': num_layers
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_norm_layer Ran
ran on a synthesized input
google-research/masksketch/masksketch/nets/layers.py
code served (permissive licence) · get_code("1983d19e3e1ab60d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned partial. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import functools
import jax
import jax.numpy as jnp
import flax.linen as nn

ARGS = {
    'train': True,
    'dtype': jnp.float32,
    'norm_type': 'BN'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
flax, jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type partial — a partial output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a partial output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
l1_loss Ran
ran on a synthesized input
google-research/masksketch/masksketch/libml/losses.py
code served (permissive licence) · get_code("06f87811f5c2b96a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ArrayImpl, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import jax.numpy as jnp
import numpy as np

# Create small sample tensors for y_true and y_pred
y_true = jnp.array([[1.0, 2.0, 3.0, 4.0],
                     [5.0, 6.0, 7.0, 8.0]], dtype=jnp.float32)

y_pred = jnp.array([[1.1, 2.1, 2.9, 4.2],
                     [4.9, 6.1, 7.1, 7.8]], dtype=jnp.float32)

ARGS = {
    "y_true": y_true,
    "y_pred": y_pred,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
restore_from_path Ran
ran on a synthesized input
google-research/masksketch/masksketch/utils.py
code served (permissive licence) · get_code("95c51644bf3cb9ff")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import flax
import tensorflow.compat.v1 as tf
import tempfile
import os

# Create a temporary directory and file for testing
temp_dir = tempfile.mkdtemp()
temp_path = os.path.join(temp_dir, "test_state.bin")

# Create a simple state object to serialize
simple_state = {"weights": [1.0, 2.0, 3.0], "bias": 0.5}

# Serialize the state using flax and write to file
serialized_bytes = flax.serialization.to_bytes(simple_state)
with tf.io.gfile.GFile(temp_path, 'wb') as f:
    f.write(serialized_bytes)

ARGS = {
    "path": temp_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
flax, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
squared_euclidean_distance Ran
ran on a synthesized input
google-research/masksketch/masksketch/libml/losses.py
code served (permissive licence) · get_code("d80134d1097de9c4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ArrayImpl, ndim 2, dtype family float, shape (2, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import jax.numpy as jnp

# Set up small test arrays
n, d, m = 2, 4, 3

# Create array `a` with shape (n, d) = (2, 4)
a = jnp.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0]
], dtype=jnp.float32)

# Create array `b` with shape (m, d) = (3, 4)
b = jnp.array([
    [1.0, 1.0, 1.0, 1.0],
    [2.0, 2.0, 2.0, 2.0],
    [3.0, 3.0, 3.0, 3.0]
], dtype=jnp.float32)

# Pre-compute b2: b square transpose with shape (1, m) = (1, 3)
b2 = jnp.sum(b.T ** 2, axis=0, keepdims=True)

# precision can be None (default) or a jax precision spec
precision = None

ARGS = {
    "a": a,
    "b": b,
    "b2": b2,
    "precision": precision
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 2, dtype family float, shape (2, 3) — shape (2, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
tensorflow_style_avg_pooling Ran
ran on a synthesized input
google-research/masksketch/masksketch/nets/layers.py
code served (permissive licence) · get_code("38e696a22dc81a6f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ArrayImpl, ndim 4, dtype family float, shape (2, 2, 1, 1) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import jax
import jax.numpy as jnp

# Set seed for reproducibility
jax.config.update('jax_platform_name', 'cpu')

# Create a small input tensor: (batch=2, height=4, width=3, channels=1)
x = jnp.ones((2, 4, 3, 1), dtype=jnp.float32)

# Define window shape for 2D pooling: (height=2, width=2)
window_shape = (2, 2)

# Define strides to match window dimensions: (stride_h=2, stride_w=2)
strides = (2, 2)

# Define padding mode
padding = 'VALID'

ARGS = {
    'x': x,
    'window_shape': window_shape,
    'strides': strides,
    'padding': padding
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 4, dtype family float, shape (2, 2, 1, 1) — shape (2, 2, 1, 1) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
truncated_normal Ran
ran on a synthesized input
google-research/masksketch/masksketch/nets/maskgit_transformer.py
code served (permissive licence) · get_code("0408240381821d2c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned function. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import jax
import jax.numpy as jnp

# Set up random key for JAX
key = jax.random.PRNGKey(0)

# Define ARGS dict with all parameters
ARGS = {
    "stddev": 0.5,
    "dtype": jnp.float32,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
upsample Ran
ran on a synthesized input
google-research/masksketch/masksketch/nets/layers.py
code served (permissive licence) · get_code("5829218c5489c6b6")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ArrayImpl, ndim 4, dtype family float, shape (2, 8, 6, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import jax
import jax.numpy as jnp

# Create a small input tensor with shape (n, h, w, c)
# Using distinct dimensions: batch=2, height=4, width=3, channels=5
x = jnp.ones((2, 4, 3, 5), dtype=jnp.float32)

# Create the ARGS dictionary mapping all parameters
ARGS = {
    'x': x,
    'factor': 2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 4, dtype family float, shape (2, 8, 6, 5) — shape (2, 8, 6, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_attn Not yet run
unverified (harvested, no recorded run)
google-research/masksketch/masksketch/sketch_conditional_inference.py
code served (permissive licence) · get_code("902758480d8399d9")
How it ran

no invocation recorded.

get_attn_vmap Not yet run
unverified (harvested, no recorded run)
google-research/masksketch/masksketch/sketch_conditional_inference.py
code served (permissive licence) · get_code("9ced563bd8005c72")
How it ran

no invocation recorded.

read_image_from_url Not yet run
unverified (harvested, no recorded run)
google-research/masksketch/masksketch/utils.py
code served (permissive licence) · get_code("e39a123f00eb58cd")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 12 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 12 later) and 3 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
get_hyper 3 other papers
1 dated earlier, 2 later
2105.12723 (2021-05)
Nested Hierarchical Transformer: Towards Accurate, Data-Efficient and Interpretable Visual Understanding
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
2311.16432 (2023-11)
Text-Driven Image Editing via Learnable Regions
apply_label_smoothing 2 other papers
1 dated earlier, 1 later
2210.00990 (2022-10)
Visual Prompt Tuning for Generative Transfer Learning
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
get_hyper 2 other papers
2 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
2311.16432 (2023-11)
Text-Driven Image Editing via Learnable Regions
tensorflow_style_avg_pooling 2 other papers
1 dated earlier, 1 later
2212.05199 (2022-12)
MAGVIT: Masked Generative Video Transformer
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
get_hyper 1 other paper
1 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
get_norm_layer 1 other paper
1 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
l1_loss 1 other paper
1 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
read_image_from_url 1 other paper
1 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
restore_from_path 1 other paper
1 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
squared_euclidean_distance 1 other paper
1 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
truncated_normal 1 other paper
1 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
upsample 1 other paper
1 later
2306.04675 (2023-06)
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Recent conditional image generation methods produce images of remarkable diversity, fidelity and realism. However, the majority of these methods allow conditioning only on labels or text prompts, which limits their level of control over the generation result. In this paper, we introduce MaskSketch, an image generation method that allows spatial conditioning of the generation result using a guiding sketch as an extra conditioning signal during sampling. MaskSketch utilizes a pre-trained masked generative transformer, requiring no model training or paired supervision, and works with input sketches of different levels of abstraction. We show that intermediate self-attention maps of a masked generative transformer encode important structural information of the input image, such as scene layout and object shape, and we propose a novel sampling method based on this observation to enable structure-guided generation. Our results show that MaskSketch achieves high image realism and fidelity to the guiding structure. Evaluated on standard benchmark datasets, MaskSketch outperforms state-of-the-art methods for sketch-to-image translation, as well as unpaired image-to-image translation approaches. The code can be found on our project website: https://masksketch.github.io/

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2302.05496")
get_code_for_paper("2302.05496")
have("2302.05496")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2302.05496.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2302.05496.svg)](https://syntology.ai/paper/2302.05496)

Connect an agent — have() is free.