SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.01482 · CVPR · 2024

EAGLE: Eigen Aggregation Learning for Object-Centric Unsupervised Semantic Segmentation

Woojung Han, Dayun Seong, Jae Hwang, Chanyoung Kim

arXiv · PDF · Open in the Atlas

Code that ran

14 of 19 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; 5 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 8 failures of our instrument; 5 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/2403.01482.json · badge: /api/ran/2403.01482.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
MICV-yonsei/EAGLE canonical (classifier label)
classifier basis: census
14 ran · 5 no recorded run
FunctionStatusWhere it lives
average_norm Ran
ran; our contract draft was wrong, not the code
MICV-yonsei/EAGLE/src_EAGLE/modules.py
code served (permissive licence) · get_code("610531000042a850")
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 Tensor, 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 torch

torch.manual_seed(0)

# The function expects a tensor t
# It performs: t / t.square().sum(1, keepdim=True).sqrt().mean()
# This requires:
# - t to have at least 2 dimensions (uses sum(1, keepdim=True))
# - t to support square(), sum(), sqrt(), mean() operations
# - Standard tensor operations

t = torch.randn(2, 4)

ARGS = {
    "t": t
}
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 Tensor, 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
bit_get Ran
ran, honoured the contract we drafted
MICV-yonsei/EAGLE/src_EAGLE/data.py
code served (permissive licence) · get_code("9c3cc207f843ec13")
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 int, ndim 0, dtype family int, 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 numpy as np

# The function bit_get takes two parameters:
# - val: an int or numpy int array
# - idx: which bit to extract
# 
# The function performs bitwise operations: val >> idx & 1
# This works on integers and numpy integer arrays

# Create test arguments
val_int = 42  # binary: 101010
idx_int = 2   # get the 2nd bit

# Also test with numpy array
val_array = np.array([15, 31, 63], dtype=np.int32)  # binary: 1111, 11111, 111111
idx_for_array = 1

ARGS = {
    "val": val_int,
    "idx": idx_int
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, 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
bool_flag Ran
ran, violated the contract we drafted
MICV-yonsei/EAGLE/src_EAGLE/dino/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from sthalles/MaSSL, is pointer only (licence: NONE) · get_code("c19231378e41dc5a")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from sthalles/MaSSL, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

clip_gradients Ran
ran, honoured the contract we drafted
MICV-yonsei/EAGLE/src_EAGLE/dino/utils.py
code served (permissive licence) · get_code("2b5489053b222e00")
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
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.linear2 = nn.Linear(3, 2)
    
    def forward(self, x):
        return self.linear2(self.linear1(x))

model = SimpleModel()

# Create a simple forward pass and backward pass to populate gradients
x = torch.randn(2, 4)
y = torch.randn(2, 2)
output = model(x)
loss = nn.MSELoss()(output, y)
loss.backward()

clip = 1.0

ARGS = {
    "model": model,
    "clip": clip,
}
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
cosine_scheduler Ran
ran, honoured the contract we drafted
MICV-yonsei/EAGLE/src_EAGLE/dino/utils.py
code served (permissive licence) · get_code("361a6b24f11fc50a")
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) and returned ndarray, ndim 1, dtype family float, shape (500,) 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 numpy as np

ARGS = {
    "base_value": 0.1,
    "final_value": 0.001,
    "epochs": 10,
    "niter_per_ep": 50,
    "warmup_epochs": 2,
    "start_warmup_value": 0.0,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (500,) — shape (500,) 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
drop_path Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
MICV-yonsei/EAGLE/src_EAGLE/dino/vision_transformer.py
pointer only (licence: MIT); the stored copy get_code serves, from gladzhang/Xformer, is pointer only (licence: NONE) · get_code("55120f2026b56aa2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from gladzhang/Xformer, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

entropy Ran
ran; our contract draft was wrong, not the code
MICV-yonsei/EAGLE/src_EAGLE/train_crf.py
code served (permissive licence) · get_code("ed616f74daccc88e")
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 Tensor, ndim 1, dtype family float, shape (2,) 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 torch

torch.manual_seed(0)

# The entropy function expects a tensor p
# It clamps minimum values, takes log, multiplies element-wise, sums over dim=1
# This suggests p should be a 2D tensor (batch_size, num_classes)
# Shape: (batch_size=2, num_classes=4)

p = torch.tensor([
    [0.1, 0.2, 0.3, 0.4],
    [0.25, 0.25, 0.25, 0.25]
], dtype=torch.float32)

ARGS = {
    "p": p
}
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 Tensor, ndim 1, dtype family float, shape (2,) — shape (2,) 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_class_labels Ran
ran; our contract draft was wrong, not the code
MICV-yonsei/EAGLE/src_EAGLE/train_segmentation_eigen.py
code served (permissive licence) · get_code("d6f15c1eec2a900c")
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

torch.manual_seed(0)

ARGS = {
    'dataset_name': 'cityscapes'
}
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
norm Ran
ran; our contract draft was wrong, not the code
MICV-yonsei/EAGLE/src_EAGLE/modules.py
pointer only (licence: MIT); the stored copy get_code serves, from alperensari/segmentation-tricks, is pointer only (licence: NONE) · get_code("12de5cf1a9536fef")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from alperensari/segmentation-tricks, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

prep_for_plot Ran
ran; our contract draft was wrong, not the code
MICV-yonsei/EAGLE/src_EAGLE/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from alperensari/segmentation-tricks, is pointer only (licence: NONE) · get_code("6a6e5c247d893c0e")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from alperensari/segmentation-tricks, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

resize Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
MICV-yonsei/EAGLE/src_EAGLE/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from alperensari/segmentation-tricks, is pointer only (licence: NONE) · get_code("d86ef5508de1f079")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from alperensari/segmentation-tricks, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

scheduler Ran
ran, honoured the contract we drafted
MICV-yonsei/EAGLE/src_EAGLE/train_segmentation_eigen.py
code served (permissive licence) · get_code("475ff955e091e303")
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 int, ndim 0, dtype family int, 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 torch
import numpy as np

class Config:
    def __init__(self, step_schedulers):
        self.step_schedulers = step_schedulers

cfg = Config(step_schedulers=10)
step = 5

ARGS = {
    "cfg": cfg,
    "step": step
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, 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
shuffle Ran
ran, honoured the contract we drafted
MICV-yonsei/EAGLE/src_EAGLE/utils.py
code served (permissive licence) · get_code("205e834237bf794c")
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 Tensor, 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 torch

torch.manual_seed(0)

x = torch.randn(2, 4)

ARGS = {
    "x": x
}
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 Tensor, 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
tensor_correlation Ran
ran; our contract draft was wrong, not the code
MICV-yonsei/EAGLE/src_EAGLE/modules.py
code served (permissive licence) · get_code("aa6344fded8d2fcc")
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 Tensor, ndim 5, dtype family float, shape (2, 3, 5, 6, 7) 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 torch

torch.manual_seed(0)

# Create two 4D tensors with shape (n, c, h, w) and (n, c, i, j)
# Using distinct dimensions: batch=2, channels=4, height=3, width=5, i=6, j=7
a = torch.randn(2, 4, 3, 5)
b = torch.randn(2, 4, 6, 7)

ARGS = {
    'a': a,
    'b': b,
}
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 Tensor, ndim 5, dtype family float, shape (2, 3, 5, 6, 7) — shape (2, 3, 5, 6, 7) 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
knn_affinity Not yet run
unverified (harvested, no recorded run)
MICV-yonsei/EAGLE/src_EAGLE/eigen_modules.py
code served (permissive licence) · get_code("40459cea1c8ce129")
How it ran

no invocation recorded.

multi_seg Not yet run
unverified (harvested, no recorded run)
MICV-yonsei/EAGLE/src_EAGLE/eigen_modules.py
code served (permissive licence) · get_code("2dcb2e9af269e7cd")
How it ran

no invocation recorded.

norm Not yet run
unverified (harvested, no recorded run)
MICV-yonsei/EAGLE/src_EAGLE/train_crf.py
code served (permissive licence) · get_code("503d86a1699da3e4")
How it ran

no invocation recorded.

prep Not yet run
unverified (harvested, no recorded run)
MICV-yonsei/EAGLE/src_EAGLE/train_crf.py
code served (permissive licence) · get_code("a6c5c2b6d88f5429")
How it ran

no invocation recorded.

rw_affinity Not yet run
unverified (harvested, no recorded run)
MICV-yonsei/EAGLE/src_EAGLE/eigen_modules.py
code served (permissive licence) · get_code("79456f0c6d9d66da")
How it ran

no invocation recorded.

Also present in other papers' code

Of 19 samples lifted from this paper's linked repositories, 19 carry a function identity and 0 do not. Of the 18 distinct functions with an identity, 12 are also present in other papers whose code we hold (11 with at least one paper dated earlier, 4 the same month, 8 later) and 6 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
drop_path 224 other papers
150 dated earlier, 6 the same month, 60 later, 7 same year, order not determinable, 1 undated · 38 only as an assembly brick
2003.12039 (2020-03)
RAFT: Recurrent All-Pairs Field Transforms for Optical Flow
2007.09384 (2020-07)
Multi-Scale Positive Sample Refinement for Few-Shot Object Detection
2010.11929 (2020-10)
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
and 221 more not named here
cosine_scheduler 37 other papers
25 dated earlier, 1 the same month, 10 later, 1 same year, order not determinable
2005.10242 (2020-05)
Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere
2103.15375 (2021-03)
AlignMixup: Improving Representations By Interpolating Aligned Features
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
and 34 more not named here
bool_flag 31 other papers
20 dated earlier, 2 the same month, 8 later, 1 same year, order not determinable
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2109.13228 (2021-09)
PASS: An ImageNet replacement for self-supervised pretraining without humans
2111.12958 (2021-11)
Self-Distilled Self-Supervised Representation Learning
and 28 more not named here
clip_gradients 28 other papers
18 dated earlier, 1 the same month, 8 later, 1 same year, order not determinable
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2109.13228 (2021-09)
PASS: An ImageNet replacement for self-supervised pretraining without humans
2111.12958 (2021-11)
Self-Distilled Self-Supervised Representation Learning
and 25 more not named here
get_class_labels 2 other papers
2 dated earlier
2203.08414 (2022-03)
Unsupervised Semantic Segmentation by Distilling Feature Correspondences
2310.17874 (2023-10)
SmooSeg: Smoothness Prior for Unsupervised Semantic Segmentation
norm 2 other papers
1 dated earlier, 1 later
2310.17874 (2023-10)
SmooSeg: Smoothness Prior for Unsupervised Semantic Segmentation
2404.03392 (2024-04)
Boosting Unsupervised Segmentation Learning
resize 2 other papers
1 dated earlier, 1 later
2203.08414 (2022-03)
Unsupervised Semantic Segmentation by Distilling Feature Correspondences
2404.03392 (2024-04)
Boosting Unsupervised Segmentation Learning
shuffle 2 other papers
1 dated earlier, 1 later
2203.08414 (2022-03)
Unsupervised Semantic Segmentation by Distilling Feature Correspondences
2404.03392 (2024-04)
Boosting Unsupervised Segmentation Learning
bit_get 1 other paper
1 dated earlier
2310.17874 (2023-10)
SmooSeg: Smoothness Prior for Unsupervised Semantic Segmentation
prep_for_plot 1 other paper
1 later
2404.03392 (2024-04)
Boosting Unsupervised Segmentation Learning
rw_affinity 1 other paper
1 dated earlier
2205.07839 (2022-05)
Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localization
tensor_correlation 1 other paper
1 dated earlier
2310.17874 (2023-10)
SmooSeg: Smoothness Prior for Unsupervised Semantic Segmentation

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

Semantic segmentation has innately relied on extensive pixel-level annotated data, leading to the emergence of unsupervised methodologies. Among them, leveraging selfsupervised Vision Transformers for unsupervised semantic segmentation (USS) has been making steady progress with expressive deep features. Yet, for semantically segmenting images with complex objects, a predominant challenge remains: the lack of explicit object-level semantic encoding in patch-level features. This technical limitation often leads to inadequate segmentation of complex objects with diverse structures. To address this gap, we present a novel approach, EAGLE, which emphasizes object-centric representation learning for unsupervised semantic segmentation. Specifically, we introduce EiCue, a spectral technique providing semantic and structural cues through an eigenbasis derived from the semantic similarity matrix of deep image features and color affinity from an image. Further, by incorporating our object-centric contrastive loss with EiCue, we guide our model to learn object-level representations with intra-and inter-image object-feature consistency, thereby enhancing semantic accuracy. Extensive experiments on COCO-Stuff, Cityscapes, and Potsdam-3 datasets demonstrate the state-of-the-art USS results of EAGLE with accurate and consistent semantic segmentation across complex scenes.

For agents

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

get_harvested_code_for_paper("2403.01482")
get_code_for_paper("2403.01482")
have("2403.01482")

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

curl https://syntology.ai/api/ran/2403.01482.json

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

[![Syntology run record](https://syntology.ai/api/ran/2403.01482.svg)](https://syntology.ai/paper/2403.01482)

Connect an agent — have() is free.