SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2404.05069 · 2024

AirShot: Efficient Few-Shot Detection for Autonomous Exploration

arXiv · PDF · Open in the Atlas

Code that ran

5 of 6 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; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 4 failures of our instrument; 1 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/2404.05069.json · badge: /api/ran/2404.05069.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
imnotprepared/airshot canonical (classifier label)
classifier basis: census
5 ran · 1 no recorded run
FunctionStatusWhere it lives
crop_support Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
imnotprepared/airshot/datasets/coco/3_gen_support_pool.py
code served (permissive licence) · get_code("549749e00d07d0be")
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 tuple. 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 numpy as np
import cv2

# Create a sample image with shape (height, width, channels)
# Using distinct dimensions: 100x80x3 to make it clear which is which
height, width, channels = 100, 80, 3
img = np.random.randint(0, 256, size=(height, width, channels), dtype=np.uint8)

# Create a bounding box [x1, y1, x2, y2]
# Bbox should be within image bounds
bbox = np.array([10, 20, 50, 60], dtype=np.float32)

ARGS = {
    "img": img,
    "bbox": bbox,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is 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
filter_coco Ran
ran; our contract draft was wrong, not the code
imnotprepared/airshot/datasets/coco/1_split_filter.py
code served (permissive licence) · get_code("b5c5db2fcaa7283b")
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 numpy as np
from collections import defaultdict

# Create a minimal mock COCO object that satisfies the function's requirements
class MockCOCO:
    def __init__(self):
        self.imgs = [1, 2, 3]  # image IDs
        self.img_data = {
            1: {'id': 1, 'file_name': 'img1.jpg'},
            2: {'id': 2, 'file_name': 'img2.jpg'},
            3: {'id': 3, 'file_name': 'img3.jpg'},
        }
        self.annotations = [
            # Image 1: annotations with category in cls_split
            {
                'id': 101,
                'image_id': 1,
                'category_id': 1,
                'bbox': [10, 10, 100, 100],  # area = 10000
                'area': 10000,
                'segmentation': [],
                'iscrowd': 0,
            },
            # Image 2: annotation with small bbox (should skip)
            {
                'id': 102,
                'image_id': 2,
                'category_id': 1,
                'bbox': [10, 10, 20, 20],  # area = 400 (< 32*32=1024)
                'area': 400,
                'segmentation': [],
                'iscrowd': 0,
            },
            # Image 3: annotation with category not in cls_split
            {
                'id': 103,
                'image_id': 3,
                'category_id': 5,
                'bbox': [10, 10, 100, 100],
                'area': 10000,
                'segmentation': [],
                'iscrowd': 0,
            },
            # Image 1 again: annotation not in cls_split
            {
                'id': 104,
                'image_id': 1,
                'category_id': 3,
                'bbox': [50, 50, 50, 50],
                'area': 2500,
                'segmentation': [],
                'iscrowd': 0,
            },
        ]
    
    def loadImgs(self, img_id):
        return [self.img_data.get(img_id, {'id': img_id})]
    
    def getAnnIds(self, imgIds, iscrowd=None):
        return [ann['id'] for ann in self.annotations if ann['image_id'] == imgIds]
    
    def loadAnns(self, ann_ids):
        return [ann for ann in self.annotations if ann['id'] in ann_ids]

# Create mock COCO instance
mock_coco = MockCOCO()

# Define cls_split: categories to filter for
cls_split = [1, 2]

ARGS = {
    'coco': mock_coco,
    'cls_split': cls_split,
}
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 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_contour_interior Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
imnotprepared/airshot/fewx/layers/boundary.py
code served (permissive licence) · get_code("37048c236eb86341")
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 tuple, ndim 3, dtype family int; 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 numpy as np
from skimage.morphology import remove_small_objects, dilation, erosion, binary_dilation, binary_erosion, square

# Create a small binary mask for testing
# Using a simple 4x5 binary array
mask = np.array([
    [0, 1, 1, 0, 1],
    [1, 1, 1, 1, 0],
    [1, 1, 0, 1, 1],
    [0, 1, 1, 1, 0]
], dtype=np.uint8)

ARGS = {
    "mask": mask,
    "bold": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, skimage — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 3, dtype family int — a tuple 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_instances_contour_interior Ran
ran, honoured the contract we drafted
imnotprepared/airshot/fewx/layers/boundary.py
code served (permissive licence) · get_code("9b506f9da113b9eb")
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 tuple, ndim 3, 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 numpy as np
import torch

# Create a simple mock object with a .data attribute
class MockInstancesMask:
    def __init__(self, data):
        self.data = data

# Create a small binary mask (4x3 for distinct dimensions)
# This represents an instances mask with some foreground regions
instances_data = np.array([
    [0, 1, 1],
    [1, 1, 0],
    [1, 1, 1],
    [0, 1, 1]
], dtype=np.uint8)

# Wrap it in the mock object that the function expects
instances_mask = MockInstancesMask(instances_data)

ARGS = {
    "instances_mask": instances_mask
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, skimage, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 3, dtype family float — a tuple 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
interpolate Ran
ran; our contract draft was wrong, not the code
imnotprepared/airshot/fewx/layers/misc.py
pointer only (licence: BSD-2-Clause); the stored copy get_code serves, from KanghoonYoon/torch-rasgg, is pointer only (licence: NONE) · get_code("2902bf4410253ff7")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (BSD-2-Clause), but get_code serves the stored copy, first harvested from KanghoonYoon/torch-rasgg, 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.

get_center Not yet run
unverified (harvested, no recorded run)
imnotprepared/airshot/fewx/layers/boundary.py
code served (permissive licence) · get_code("dac8413f65a84dfc")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 4 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 1 later) and 2 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
interpolate 38 other papers
33 dated earlier, 5 later
1703.06870 (2017-03)
Mask R-CNN
1803.03243 (2018-03)
Domain Adaptive Faster R-CNN for Object Detection in the Wild
1807.08865 (2018-07)
StereoNet: Guided Hierarchical Refinement for Real-Time Edge-Aware Depth Prediction
and 35 more not named here
get_center 3 other papers
3 dated earlier
1908.01998 (2019-08)
Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector
2103.12340 (2021-03)
Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers
136990421 (2022 (year only))
AirDet: Few-Shot Detection without Fine-tuning for Autonomous Exploration
get_contour_interior 3 other papers
3 dated earlier
1908.01998 (2019-08)
Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector
2103.12340 (2021-03)
Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers
136990421 (2022 (year only))
AirDet: Few-Shot Detection without Fine-tuning for Autonomous Exploration
get_instances_contour_interior 3 other papers
3 dated earlier
1908.01998 (2019-08)
Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector
2103.12340 (2021-03)
Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers
136990421 (2022 (year only))
AirDet: Few-Shot Detection without Fine-tuning for Autonomous Exploration

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

Few-shot object detection has drawn increasing attention in the field of robotic exploration, where robots are required to find unseen objects with a few online provided examples. Despite recent efforts have been made to yield online processing capabilities, slow inference speeds of low-powered robots fail to meet the demands of real-time detection-making them impractical for autonomous exploration. Existing methods still face performance and efficiency challenges, mainly due to unreliable features and exhaustive class loops. In this work, we propose a new paradigm AirShot, and discover that, by fully exploiting the valuable correlation map, AirShot can result in a more robust and faster few-shot object detection system, which is more applicable to robotics community. The core module Top Prediction Filter (TPF) can operate on multi-scale correlation maps in both the training and inference stages. During training, TPF supervises the generation of a more representative correlation map, while during inference, it reduces looping iterations by selecting top-ranked classes, thus cutting down on computational costs with better performance. Surprisingly, this dual functionality exhibits general effectiveness and efficiency on various off-the-shelf models. Exhaustive experiments on COCO2017, VOC2014, and SubT datasets demonstrate that TPF can significantly boost the efficacy and efficiency of most off-the-shelf models, achieving up to 36.4% precision improvements along with 56.3% faster inference speed. Code and Data are at: https://github.com/ImNotPrepared/AirShot.

For agents

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

get_harvested_code_for_paper("2404.05069")
get_code_for_paper("2404.05069")
have("2404.05069")

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

curl https://syntology.ai/api/ran/2404.05069.json

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

[![Syntology run record](https://syntology.ai/api/ran/2404.05069.svg)](https://syntology.ai/paper/2404.05069)

Connect an agent — have() is free.