RGBT Tracking via All-layer Multimodal Interactions with Progressive Fusion Mamba
Authors: Andong Lu, Wanyu Wang, Chenglong Li, Jin Tang, Bin Luo
AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on four public RGBT tracking datasets show that AINet achieves leading performance against existing stateof-the-art methods. ... We evaluate our proposed AINet on four popular RGBT tracking benchmarks: RGBT210, RGBT234, Las He R and VTUAV, and compare the performance with 20 state-of-the-art RGBT trackers. We adopt the Precision Rate (PR), Success Rate (SR), and Normalized Precision Rate (NPR) from One-Pass Evaluation (OPE) as metrics for quantitative performance measurement... In Table 2, we conduct several ablation studies on the Las He R dataset to verify the effectiveness of key components in our AINet. |
| Researcher Affiliation | Academia | 1School of Computer Science and Technology, Anhui University 2School of Artificial Intelligence, Anhui University adlu EMAIL, EMAIL, EMAIL, EMAIL, EMAIL |
| Pseudocode | No | The paper describes the architecture and methodologies (DFM, OFM) using equations and textual descriptions, but it does not include any explicitly labeled pseudocode blocks or algorithms. |
| Open Source Code | No | We will release the code upon acceptance of the paper. |
| Open Datasets | Yes | Extensive experiments on four public RGBT tracking datasets show that AINet achieves leading performance against existing stateof-the-art methods. ...We evaluate our proposed AINet on four popular RGBT tracking benchmarks: RGBT210, RGBT234, Las He R and VTUAV |
| Dataset Splits | Yes | We use the Las He R training set to train our network, which is used to evaluate RGBT210 (Li et al. 2017), RGB234 (Li et al. 2019) and Las He R (Li et al. 2021). For the evaluation of VTUAV (Zhang et al. 2022), we utilize the training set from VTUAV as the training data. |
| Hardware Specification | Yes | We implement our AINet based on the Py Torch and train it on single NVIDIA RTX 4090 GPU. |
| Software Dependencies | No | We implement our AINet based on the Py Torch and train it on single NVIDIA RTX 4090 GPU. The paper mentions PyTorch but does not specify a version number or other software dependencies with versions. |
| Experiment Setup | Yes | We follow the hyperparameter settings of the baseline model (Ye et al. 2022) for the loss function. For parameter initialization, we utilize the pretrained model provided by Drop Track (Wu et al. 2023). During training, we use the Adam W (Loshchilov and Hutter 2017) optimizer with a weight decay of 10 4, and set the batch size and learning rate to 16 and 10 4, respectively. The entire network is trained end-to-end over 15 epochs, with each epoch providing 6 104 pairs of samples. |