Papers › Instruction-Guided Visual Masking

Instruction-Guided Visual Masking

30 May 2024arXiv:2405.19783archive 2025-07-28

Jinliang Zheng, Jianxiong Li, Sijie Cheng, Yinan Zheng, Jiaming Li, Jihao Liu, Yu Liu, Jingjing Liu, Xianyuan Zhan

Instruction following is crucial in contemporary LLM. However, when extended to multimodal setting, it often suffers from misalignment between specific textual instruction and targeted local region of an image. To achieve more accurate and nuanced multimodal instruction following, we introduce Instruction-guided Visual Masking (IVM), a new versatile visual grounding model that is compatible with diverse multimodal models, such as LMM and robot model. By constructing visual masks for instruction-irrelevant regions, IVM-enhanced multimodal models can effectively focus on task-relevant image regions to better align with complex instructions. Specifically, we design a visual masking data generation pipeline and create an IVM-Mix-1M dataset with 1 million image-instruction pairs. We further introduce a new learning technique, Discriminator Weighted Supervised Learning (DWSL) for preferential IVM training that prioritizes high-quality data samples. Experimental results on generic multimodal tasks such as VQA and embodied robotic control demonstrate the versatility of IVM, which as a plug-and-play tool, significantly boosts the performance of diverse multimodal models, yielding new state-of-the-art results across challenging multimodal benchmarks. Code, model and data are available at https://github.com/2toinf/IVM.

PaperPDFCodeCode Syntology ranMethod as code

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 28 code samples harvested from 1 repository linked to this paper: 20 ran (of which 7 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 10 where Syntology's instrument failed) · 8 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; 9 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 28 samples from 1 repository, 20 ran (10 with no instrument failure; 10 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 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 20 samples that ran has no recorded date.

2toinf/ivm officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: 28 samples, 20 ran (10 instrument-failed) · 8 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

28 samples harvested: 20 ran (of which 7 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 10 where Syntology's instrument failed) · 8 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
9ran; the contract check could not run
9ran
8unverified

Licence: 2 of the 28 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 2toinf/ivm. “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”. For 4 samples that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “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. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

2toinf/ivm official repository 28 samples, 20 ran (10 instrument-failed) · 8 unverified
Attention in 344 papers
census: 13 implementations from 12 papers, 10 distinct outputs across 4 buckets, one shared input per bucket
model/IVM.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · b5bf903edd1dcafd · report fixture: returned Attention · level recorded 2026-09-22
ConditionedMaskDecoder model/IVM.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 256eb3b402dbb02a · report fixture: returned ConditionedMaskDecoder · level recorded 2026-09-22
LayerNorm2d in 18 papers model/IVM.py ran · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · 738b5dcde12052ba · report fixture: returned LayerNorm2d · level recorded 2026-09-22
MLP in 375 papers model/IVM.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · d0650d05ab2c4a12 · report fixture: returned MLP · level recorded 2026-09-22
MLPBlock in 17 papers model/IVM.py ran · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · 1da9d87f536d9241 · report fixture: returned MLPBlock · level recorded 2026-09-22
PatchEmbed in 123 papers
census: 9 implementations from 9 papers, 6 distinct outputs on one shared input
model/IVM.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 7ed2e6b526e58fd3 · report fixture: returned PatchEmbed · level recorded 2026-09-22
PositionEmbeddingRandom in 8 papers model/IVM.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 2cb4a7bc53ea5022 · report fixture: returned PositionEmbeddingRandom · level recorded 2026-09-22
add_decomposed_rel_pos in 7 papers model/IVM.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant Apache-2.0 (permissive) · 30cb45bcc9fc27d0 · report fixture: signature attn:Tensor/3/float, q:Tensor/3/float, rel_pos_h:Tensor/2/float, rel_pos_w:Tensor/2/float, q_size:tuple/1/int, k_size:tuple/1/int · returned Tensor · level recorded 2026-09-22
apply_mask_to_rgb model/llava_encoder.py ran Apache-2.0 (permissive) · 3a06fee7a1f09df9 · report fixture: signature image:Image/0/?, mask:Tensor/2/float, fill_color:tuple/1/int · returned Image · level recorded 2026-09-25
dice_loss in 193 papers
census: 10 implementations from 16 papers, 7 distinct outputs across 4 buckets, one shared input per bucket this sample: bucket 3 of 4, arg 1: rank 4, kind float, dtype float32 · arg 2: rank 4, kind float, dtype float32
model/IVM.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted Apache-2.0 (permissive) · d91fd9b31c41684e · report fixture: signature inputs:Tensor/4/float, targets:Tensor/4/float, scale:int/0/int, eps:float/0/float · returned Tensor · level recorded 2026-09-22
generate_boxes_mask utils.py ran Apache-2.0 (permissive) · 33b1f0add9c27a54 · report fixture: signature xmin:float/0/float, ymin:float/0/float, xmax:float/0/float, ymax:float/0/float, height:int/0/int, width:int/0/int · returned ndarray · level recorded 2026-09-25
get_rel_pos in 54 papers
census: 3 implementations from 6 papers, 3 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
model/sam_components/image_encoder.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant Apache-2.0 (permissive) · 733d7f0bedcb74c2 · report fixture: signature q_size:int/0/int, k_size:int/0/int, rel_pos:Tensor/2/float · returned Tensor · level recorded, date not recorded
get_rel_pos in 54 papers
census: 3 implementations from 6 papers, 3 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
model/IVM.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant Apache-2.0 (permissive) · 5cf5ba2c32195b02 · report fixture: signature q_size:int/0/int, k_size:int/0/int, rel_pos:Tensor/2/float · returned Tensor · level recorded 2026-09-22
llavaTextPreprocess model/IVM.py ran · our draft was wrong Apache-2.0 (permissive) · eeb6e13ac80aed43 · report fixture: signature ins:str/0/str, role:str/0/str · returned str · level recorded 2026-09-22
read_masks IVM_1M_Mix/datareader.py ran · honoured contract Apache-2.0 (permissive) · 7e8dcee39799a2b0 · report fixture: signature mask_path:str/0/str · returned ndarray · level recorded 2026-09-25
sigmoid_ce_loss in 70 papers
census: 4 implementations from 8 papers, 3 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 3, kind float, dtype float32 · arg 2: rank 3, kind float, dtype float32
model/IVM.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted Apache-2.0 (permissive) · 1033e9116b1e2573 · report fixture: signature inputs:Tensor/3/float, targets:Tensor/3/float · returned Tensor · level recorded 2026-09-22
window_partition in 377 papers
census: 13 implementations from 17 papers, 5 distinct outputs across 3 buckets, one shared input per bucket
model/IVM.py ran; the contract check could not run (RAISES) · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · 89c1f2e801f0e941 · report fixture: signature x:Tensor/4/float, window_size:int/0/int · returned tuple · level recorded 2026-09-22
window_partition in 377 papers
census: 13 implementations from 17 papers, 5 distinct outputs across 3 buckets, one shared input per bucket
model/sam_components/image_encoder.py ran; the contract check could not run (RAISES) · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) here; the stored copy is pointer only · 105fa08885dc36cc · report fixture: signature x:Tensor/4/float, window_size:int/0/int · returned tuple · level recorded 2026-09-22
window_unpartition in 56 papers model/IVM.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant Apache-2.0 (permissive) · 3ade8cb3a2133a3f · report fixture: signature windows:Tensor/4/float, window_size:int/0/int, pad_hw:tuple/1/int, hw:tuple/1/int · returned Tensor · level recorded 2026-09-22
window_unpartition in 56 papers model/sam_components/image_encoder.py ran; the contract check could not run (RAISES) Apache-2.0 (permissive) here; the stored copy is pointer only · 27be441cc8213e52 · report fixture: signature windows:Tensor/4/float, window_size:int/0/int, pad_hw:tuple/1/int, hw:tuple/1/int · returned Tensor · level recorded 2026-09-26
IVM model/IVM.py unverified Apache-2.0 (permissive) · 28446a46dff45e88 · report
ImageEncoderViT in 7 papers model/IVM.py unverified Apache-2.0 (permissive) · d930ca853fbfe859 · report
LLaVAEncoder model/IVM.py unverified Apache-2.0 (permissive) · d6c5069b5e0fd935 · report
TwoWayAttentionBlock model/IVM.py unverified Apache-2.0 (permissive) · 718ac29e8d9232bb · report
TwoWayTransformer model/IVM.py unverified Apache-2.0 (permissive) · 227d8c19ab9e7cbb · report
auto_postprocess IVM.py unverified Apache-2.0 (permissive) · 58ede5faa9cd2f1c · report
forward_batch IVM.py unverified Apache-2.0 (permissive) · 1f0e4db7df0cd3fe · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for Instruction-guided Visual Masking (IVM) (graph method: Instruction-guided Visual Masking (IVM)). 20 bricks; individually on a fixture: 14 ran (of which 7 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 7 where Syntology's instrument failed) · 6 unverified. The assembly has not been run end-to-end.

  1. IVM entry · level 0 · unverified · 28446a46dff45e88
  2. ConditionedMaskDecoder model · level 1 · ran · 256eb3b402dbb02a
  3. ImageEncoderViT helper · level 0 · unverified · d930ca853fbfe859
  4. LLaVAEncoder helper · level 0 · unverified · d6c5069b5e0fd935
  5. TwoWayTransformer helper · level 0 · unverified · 227d8c19ab9e7cbb
  6. dice_loss loss · level 1 · ran; the contract check could not run (RAISES) · d91fd9b31c41684e
  7. sigmoid_ce_loss loss · level 1 · ran; the contract check could not run (RAISES) · 1033e9116b1e2573
  8. LayerNorm2d helper · level 1 · ran · 738b5dcde12052ba
  9. MLP helper · level 1 · ran · d0650d05ab2c4a12
  10. PositionEmbeddingRandom helper · level 1 · ran · 2cb4a7bc53ea5022
  11. Block helper · level 0 · unverified · 2dcf225b97d045ae
  12. PatchEmbed helper · level 1 · ran · 7ed2e6b526e58fd3
  13. llavaTextPreprocess helper · level 1 · ran · our draft was wrong · eeb6e13ac80aed43
  14. Attention helper · level 1 · ran · b5bf903edd1dcafd
  15. TwoWayAttentionBlock helper · level 0 · unverified · 718ac29e8d9232bb
  16. MLPBlock helper · level 1 · ran · 1da9d87f536d9241
  17. window_partition helper · level 1 · ran; the contract check could not run (RAISES) · 89c1f2e801f0e941
  18. window_unpartition helper · level 1 · ran; the contract check could not run (RAISES) · 3ade8cb3a2133a3f
  19. add_decomposed_rel_pos helper · level 1 · ran; the contract check could not run (RAISES) · 30cb45bcc9fc27d0
  20. get_rel_pos helper · level 1 · ran; the contract check could not run (RAISES) · 5cf5ba2c32195b02

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2405.19783") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Instruction FollowingVisual GroundingVisual Question AnsweringVisual Question Answering (VQA)

Datasets

Introduced by this paper, per the archive.

IVM-Mix-1M

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Visual Question Answering V*bench IVM-Enhanced GPT4-V Accuracy 81.2 #3 of 5 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

ALIGNFocus

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