Papers › Phantom: Subject-consistent video generation via cross-modal alignment

Phantom: Subject-consistent video generation via cross-modal alignment

16 Feb 2025arXiv:2502.11079archive 2025-07-28

Lijie Liu, Tianxiang Ma, Bingchuan Li, Zhuowei Chen, Jiawei Liu, Qian He, Xinglong Wu

The continuous development of foundational models for video generation is evolving into various applications, with subject-consistent video generation still in the exploratory stage. We refer to this as Subject-to-Video, which extracts subject elements from reference images and generates subject-consistent video through textual instructions. We believe that the essence of subject-to-video lies in balancing the dual-modal prompts of text and image, thereby deeply and simultaneously aligning both text and visual content. To this end, we propose Phantom, a unified video generation framework for both single and multi-subject references. Building on existing text-to-video and image-to-video architectures, we redesign the joint text-image injection model and drive it to learn cross-modal alignment via text-image-video triplet data. In particular, we emphasize subject consistency in human generation, covering existing ID-preserving video generation while offering enhanced advantages. The project homepage is here https://phantom-video.github.io/Phantom/.

PaperPDFCodeCode Syntology ran

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

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2502.11079")

Code

Syntology Ran 3 of 12 code samples harvested from 1 repository linked to this paper; 9 have no recorded run. Of those that ran: 2 ran · our draft was wrong; 1 ran · fixture could not drive it.

By repository: community (archive-listed): 12 samples from 1 repository, 3 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

phantom-video/phantom mentioned on GitHubpytorchApache-2.0 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

12 samples harvested; 3 ran; 0 honoured the contract we drafted; 9 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

2ran · our draft was wrong
1ran · fixture could not drive it
9unverified

Licence: 0 of the 12 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from phantom-video/phantom. “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. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (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 from a repository marked official are listed first. 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 four 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; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). 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.

basic_clean phantom-video/phantom/phantom_wan/modules/tokenizers.py community (archive-listed) ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 98f385d847636a3e · report
sinusoidal_embedding_1d phantom-video/phantom/phantom_wan/modules/model.py community (archive-listed) ran · fixture could not drive it fingerprinted Apache-2.0 (permissive) · e34a97853e924d88 · report
whitespace_clean phantom-video/phantom/phantom_wan/modules/tokenizers.py community (archive-listed) ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 9542161e9640b858 · report
attention phantom-video/phantom/phantom_wan/modules/attention.py community (archive-listed) unverified Apache-2.0 (permissive) · 2659ec11ef2cea0d · report
canonicalize phantom-video/phantom/phantom_wan/modules/tokenizers.py community (archive-listed) unverified Apache-2.0 (permissive) · 1066f2140d4e9845 · report
count_conv3d phantom-video/phantom/phantom_wan/modules/vae.py community (archive-listed) unverified Apache-2.0 (permissive) · 78d2ce0959095290 · report
flash_attention phantom-video/phantom/phantom_wan/modules/attention.py community (archive-listed) unverified Apache-2.0 (permissive) · 7ea44a4680b76539 · report
fp16_clamp phantom-video/phantom/phantom_wan/modules/t5.py community (archive-listed) unverified Apache-2.0 (permissive) · d3d53f877a9af3a3 · report
pos_interpolate phantom-video/phantom/phantom_wan/modules/clip.py community (archive-listed) unverified Apache-2.0 (permissive) · cbc3565ced3e12b6 · report
rope_apply phantom-video/phantom/phantom_wan/modules/model.py community (archive-listed) unverified Apache-2.0 (permissive) · 3571f7fc8c1b5da7 · report
rope_params phantom-video/phantom/phantom_wan/modules/model.py community (archive-listed) unverified Apache-2.0 (permissive) · b81485ecf664d0d5 · report
xlm_roberta_large phantom-video/phantom/phantom_wan/modules/xlm_roberta.py community (archive-listed) unverified Apache-2.0 (permissive) · dd27a4bda5e8824c · report

Tasks

Human-Domain Subject-to-VideoOpen-Domain Subject-to-VideoSingle-Domain Subject-to-VideoVideo Generationcross-modal alignment

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-14B Aesthetics 0.4639 #3 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-14B FaceSim 0.5148 #3 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-14B GmeScore 0.7065 #3 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-14B Motion 0.3342 #3 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-14B NaturalScore 0.6866 #3 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-14B NexusScore 0.3743 #3 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-14B Total Score 0.5232 #3 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-14B Venue Open-Source #3 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-1.3B Aesthetics 0.4667 #4 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-1.3B FaceSim 0.4855 #4 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-1.3B GmeScore 0.6942 #4 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-1.3B Motion 0.1429 #4 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-1.3B NaturalScore 0.7026 #4 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-1.3B NexusScore 0.4244 #4 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-1.3B Total Score 0.5071 #4 of 10 Archive leaderboard report
Open-Domain Subject-to-Video OpenS2V-Eval Phantom-Wan-1.3B Venue Open-Source #4 of 10 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.

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