Papers › Vision-by-Language for Training-Free Compositional Image Retrieval

Vision-by-Language for Training-Free Compositional Image Retrieval

13 Oct 2023arXiv:2310.09291archive 2025-07-28

Shyamgopal Karthik, Karsten Roth, Massimiliano Mancini, Zeynep Akata

Given an image and a target modification (e.g an image of the Eiffel tower and the text "without people and at night-time"), Compositional Image Retrieval (CIR) aims to retrieve the relevant target image in a database. While supervised approaches rely on annotating triplets that is costly (i.e. query image, textual modification, and target image), recent research sidesteps this need by using large-scale vision-language models (VLMs), performing Zero-Shot CIR (ZS-CIR). However, state-of-the-art approaches in ZS-CIR still require training task-specific, customized models over large amounts of image-text pairs. In this work, we propose to tackle CIR in a training-free manner via our Compositional Image Retrieval through Vision-by-Language (CIReVL), a simple, yet human-understandable and scalable pipeline that effectively recombines large-scale VLMs with large language models (LLMs). By captioning the reference image using a pre-trained generative VLM and asking a LLM to recompose the caption based on the textual target modification for subsequent retrieval via e.g. CLIP, we achieve modular language reasoning. In four ZS-CIR benchmarks, we find competitive, in-part state-of-the-art performance - improving over supervised methods. Moreover, the modularity of CIReVL offers simple scalability without re-training, allowing us to both investigate scaling laws and bottlenecks for ZS-CIR while easily scaling up to in parts more than double of previously reported results. Finally, we show that CIReVL makes CIR human-understandable by composing image and text in a modular fashion in the language domain, thereby making it intervenable, allowing to post-hoc re-align failure cases. Code will be released upon acceptance.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 8 code samples harvested from 1 repository linked to this paper: 4 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 1 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 4 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 8 samples from 1 repository, 4 ran (3 with no instrument failure; 1 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

explainableml/vision_by_language officialmentioned in papermentioned on GitHubpytorchMITSyntology: 8 samples, 4 ran (1 instrument-failed) · 4 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

8 samples harvested: 4 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 1 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 4 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
1ran · our draft was wrong
2ran
4unverified

Licence: 2 of the 8 samples are pointer only, meaning Syntology does not serve that copy's text. For 2 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from explainableml/vision_by_language. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

explainableml/vision_by_language official repository 8 samples, 4 ran (1 instrument-failed) · 4 unverified
collate_fn in 500 papers src/data_utils.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only · b6c1f837e6faaeba · report fixture: signature batch:list/1/tuple · returned list · level recorded 2026-09-22
expand2square in 78 papers src/datasets.py ran MIT (permissive) · e75146036d2d1504 · report fixture: signature pil_img:Image/0/?, background_color:tuple/1/int · returned Image · level recorded 2026-09-24
get_recall in 7 papers src/utils.py ran MIT (permissive) · 55917bc288f17d17 · report fixture: signature indices:Tensor/2/int, targets:Tensor/1/int · returned float · level recorded 2026-09-24
targetpad_transform in 11 papers src/data_utils.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · 9476d6cebb3ecfb9 · report fixture: signature target_ratio:float/0/float, dim:int/0/int · returned Compose · level recorded 2026-09-24
circo src/compute_results.py unverified MIT (permissive) · afe61f6410fb50d2 · report
cirr src/compute_results.py unverified MIT (permissive) · 0cc283db93cbae1d · report
fiq src/compute_results.py unverified MIT (permissive) · 5ee1f8a4e8bf641d · report
openai_completion in 7 papers src/openai_api.py unverified MIT (permissive) · f6c8fb5e60546cec · report

Tasks

Image RetrievalRetrievalZero-Shot Composed Image Retrieval (ZS-CIR)

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRCO CIReVL (CLIP G/14) mAP@10 27.59 #15 of 43 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRCO CIReVL (CLIP L/14) mAP@10 19.01 #26 of 43 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRCO CIReVL (CLIP B/32) mAP@10 15.42 #30 of 43 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRR CIReVL (CLIP G/14) R@1 34.65 #10 of 47 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRR CIReVL (CLIP G/14) R@5 64.29 #10 of 47 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRR CIReVL (CLIP L/14) R@1 24.55 #21 of 47 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRR CIReVL (CLIP L/14) R@5 52.31 #21 of 47 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRR CIReVL (CLIP B/32) R@1 23.94 #22 of 47 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) CIRR CIReVL (CLIP B/32) R@5 52.51 #22 of 47 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) Fashion IQ CIReVL (CLIP G/14) (Recall@10+Recall@50)/2 42.28 #20 of 41 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) Fashion IQ CIReVL (CLIP B/32) (Recall@10+Recall@50)/2 38.82 #27 of 41 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) Fashion IQ CIReVL (CLIP L/14) (Recall@10+Recall@50)/2 38.56 #28 of 41 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) GeneCIS CIReVL (CLIP B/32) A-R@1 15.9 #9 of 11 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) GeneCIS CIReVL (CLIP L/14) A-R@1 15.9 #10 of 11 Archive leaderboard report
Zero-Shot Composed Image Retrieval (ZS-CIR) GeneCIS CIReVL (CLIP G/14) A-R@1 17.4 #11 of 11 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

BLIPCLIP

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections