Transparent Object Tracking Benchmark

Heng Fan     Halady Akhilesha Miththanthaya*     Harshit*     Siranjiv Ramana Rajan*    
Xiaoqiong Liu     Zhilin Zou     Yuewei Lin     Haibin Ling

University of North Texas        Stony Brook University        Brookhaven National Laboratory


Visual tracking has achieved considerable progress in recent years. However, current research in the field mainly focuses on tracking of opaque objects, while little attention is paid to transparent object tracking. In this paper, we make the first attempt in exploring this problem by proposing a Transparent Object Tracking Benchmark (TOTB). Specifically, TOTB consists of 225 videos (86K frames) from 15 diverse transparent object categories. Each sequence is manually labeled with axis-aligned bounding boxes. To the best of our knowledge, TOTB is the first benchmark dedicated to transparent object tracking. In order to understand how existing trackers perform and to provide comparison for future research on TOTB, we extensively evaluate 25 state-of-the-art tracking algorithms. The evaluation results exhibit that more efforts are needed to improve transparent object tracking. Besides, we observe some nontrivial findings from the evaluation that are discrepant with some common beliefs in opaque object tracking. For example, we find that deeper features are not always good for improvements. Moreover, to encourage future research, we introduce a novel tracker, named TransATOM, which leverages transparency features for tracking and surpasses all 25 evaluated approaches by a large margin. By releasing TOTB, we expect to facilitate future research and application of transparent object tracking in both the academia and industry.


225 Videos (Targets)        ~86,000 Frames        15 Transparent Object Categories        12 Attributes

Qualitative examples of some of the video sequences contained in TOTB.

A New Baseline: TransATOM

Comparison between the state-of-the-art ATOM and the proposed new baseline TransATOM which integrates conventional classification feature and transparency feature (check details in the paper) for better target localization. Both methods use the same IoUNet for target scale estimation.

Classification results of ATOM and TransATOM. We can observe that TransATOM shows better classification results for locating transparent target objects. The yellow boxes in input images are groundtruth.


Tracking performance of 25 state-of-the-art trackers and TransATOM on TOTB using precision, normalized precision and success.

Tracking performance of different trackers on the three most common attributes in TOTB including ROT, POC and SV.

Qualitative tracking results of state-of-the-art trackers and TransATOM on TOTB.


Paper     Dataset (google drive~7.8G)     Evaluation toolkit     TransATOM code


        title={Transparent Object Tracking Benchmark},
        author={Fan, Heng and Miththanthaya, Halady Akhilesha and Harshit and Rajan,
        Siranjiv Ramana and Liu, Xiaoqiong and Zou, Zhilin and Lin, Yuewei and Ling, Haibin},
        booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},

Contact: If you have any questions, please contact Heng Fan at