SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.16002 · CVPR · 2024

SDSTrack: Self-Distillation Symmetric Adapter Learning for Multi-Modal Visual Object Tracking

Zhengkai Jiang, Yong Liu, Liang Liu, Junhao Chen, Mengmeng Wang, Jiazheng Xing, Yijie Qian, Kai Tang, Xiaojun Hou, Yaowei Guo, Shuo Xin

arXiv · PDF · Open in the Atlas

Code that ran

1 of 1 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 0 failures of our instrument; 0 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/2403.16002.json · badge: /api/ran/2403.16002.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
hoqolo/SDSTrack community
Papers with Code archive
1 ran · 0 no recorded run
FunctionStatusWhere it lives
SDSTrack Ran
ran on a synthesized input · constructed an object; no result was computed
hoqolo/SDSTrack/lib/models/sdstrack/sdstrack.py
code served (permissive licence) · get_code("4e59c211afbdecef")
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 an instance of its own class, SDSTrack: the run constructed the object and did not call it, so no result was computed. 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 torch
from torch import nn

torch.manual_seed(0)

class DummyTransformer(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 4)
    
    def forward(self, z, x, ce_template_mask=None, ce_keep_rate=None, return_last_attn=False, infer=False):
        # z: template, x: search
        # Return x_list and aux_dict_list
        # x_list should be list of tensors or list of lists of tensors
        # Each element should have shape where last feat_len_s elements are used
        
        # feat_sz_s = 2, so feat_len_s = 4
        # We need cat_feature of shape (HW1+HW2, B, C) or (HW2, B, C)
        # After slicing [:, -4:] we get (something, B, 4)
        # Then unsqueeze(-1) -> (something, B, 4, 1)
        # permute (0,3,2,1) -> (something, 1, 4, B)
        # Then unsqueeze and view to get (B, Nq, C, HW) where HW=4
        
        # So we need output where -4 elements give us what we need
        # Let's make shape (10, 2, 4) where last 4 elements are the important ones
        feat = torch.randn(10, 2, 4)
        x_list = [feat]
        aux_dict_list = [{}]
        return x_list, aux_dict_list

class DummyBoxHead(nn.Module):
    def __init__(self, feat_sz=2):
        super().__init__()
        self.feat_sz = feat_sz
        self.conv = nn.Conv2d(4, 4, 1)
    
    def forward(self, opt_feat, arg2=None):
        # opt_feat shape: (B, C, feat_sz_s, feat_sz_s) = (B, 4, 2, 2)
        # Need to return pred_box and score_map for CORNER head
        # pred_box should be (B*Nq, 4) in xyxy format
        # score_map can be anything
        
        bs = opt_feat.shape[0]
        pred_box = torch.randn(bs, 4)
        score_map = torch.randn(bs, 1, 2, 2)
        return pred_box, score_map

class DummyBoxHeadCenter(nn.Module):
    def __init__(self, feat_sz=2):
        super().__init__()
        self.feat_sz = feat_sz
        self.conv = nn.Conv2d(4, 4, 1)
    
    def forward(self, opt_feat, gt_score_map=None):
        # For CENTER head: returns score_map_ctr, bbox, size_map, offset_map
        bs = opt_feat.shape[0]
        score_map_ctr = torch.randn(bs, 1, 2, 2)
        bbox = torch.randn(bs, 4)
        size_map = torch.randn(bs, 2, 2, 2)
        offset_map = torch.randn(bs, 2, 2, 2)
        return score_map_ctr, bbox, size_map, offset_map

transformer = DummyTransformer()
box_head = DummyBoxHead(feat_sz=2)

ARGS = {
    'transformer': transformer,
    'box_head': box_head,
    'aux_loss': False,
    'head_type': 'CORNER'
}
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 SDSTrack — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a SDSTrack output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not 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

Also present in other papers' code

Of 1 samples lifted from this paper's linked repositories, 1 carry a function identity and 0 do not. Of the 1 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 1 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.

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

Multimodal Visual Object Tracking (VOT) has recently gained significant attention due to its robustness. Early research focused on fully fine-tuning RGB-based trackers, which was inefficient and lacked generalized representation due to the scarcity of multimodal data. Therefore, recent studies have utilized prompt tuning to transfer pretrained RGB-based trackers to multimodal data. However, the modality gap limits pre-trained knowledge recall, and the dominance of the RGB modality persists, preventing the full utilization of information from other modalities. To address these issues, we propose a novel symmetric multimodal tracking framework called SDSTrack. We introduce lightweight adaptation for efficient fine-tuning, which directly transfers the feature extraction ability from RGB to other domains with a small number of trainable parameters and integrates multimodal features in a balanced, symmetric manner. Furthermore, we design a complementary masked patch distillation strategy to enhance the robustness of trackers in complex environments, such as extreme weather, poor imaging, and sensor failure. Extensive experiments demonstrate that SDSTrack outperforms state-ofthe-art methods in various multimodal tracking scenarios, including RGB+Depth, RGB+Thermal, and RGB+Event tracking, and exhibits impressive results in extreme conditions. Our source code is available at: https:// github.com/hoqolo/SDSTrack.

For agents

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

get_harvested_code_for_paper("2403.16002")
get_code_for_paper("2403.16002")
have("2403.16002")

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

curl https://syntology.ai/api/ran/2403.16002.json

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

[![Syntology run record](https://syntology.ai/api/ran/2403.16002.svg)](https://syntology.ai/paper/2403.16002)

Connect an agent — have() is free.