Monocular Model-Based 3D Object Tracking Methods



** Update: July 2026. Please send me your results via email (byungkuk.seo@etri.re.kr) for the lastest updates!

## RBOT Dataset (https://github.com/henningtjaden/RBOT)

Appearance Ape Soda Vise Soup Camera Can Cat Clown Cube Driller Duck Egg Glue Iron Candy Lamp Phone Squirrel
[H. Tjaden et al.]
TPAMI (2019)
well-textured
weakly-textured
[X. Sun et al.]
TCSVT (2021)
very-distinct
less-distinct

## Performance Comparison (SR (%))

The best results are highlighted as First, Second, and Third.

Regular scene

<
Method Ape Soda Vise Soup Camera Can Cat Clown Cube Driller Duck Egg Glue Iron Candy Lamp Phone Squirrel Avg.
G. Wang et al. (a) VC (2015) 13.3 13.3 16.1 16.0 15.1 14.6 19.3 14.0 13.9 16.6 16.5 11.0 13.4 15.6 8.7 15.9 13.5 18.2 14.7
G. Wang et al. (b) VC (2015) 5.5 3.9 5.9 7.9 5.5 4.7 16.8 3.6 3.8 6.2 7.2 2.3 2.8 5.8 1.2 3.8 4.6 12.9 5.8
H. Tjaden et al. (c) ICCV (2017) 62.1 30.5 95.8 66.2 61.6 81.7 96.7 89.1 44.1 87.7 74.9 50.9 20.2 68.4 20.0 92.3 64.9 98.5 67.0
B. Wang et al. (d) MTAP (2019) 22.4 19.4 22.9 21.3 23.3 19.8 20.7 23.1 22.7 22.4 20.5 21.2 19.7 22.2 21.0 23.0 24.9 22.6 21.8
L. Zhong et al. (e) IJCV (2019) 82.6 40.1 92.6 85.0 82.8 87.2 98.0 92.9 81.3 84.5 83.3 76.2 56.1 84.6 57.6 90.5 82.6 95.6 80.8
H. Tjaden et al. TPAMI (2019) 85.0 39.0 98.9 82.4 79.9 87.6 95.9 93.3 78.1 93.0 86.8 74.6 38.9 81.0 46.8 97.5 80.7 99.4 79.9
H. Huang et al. CGF (2020) 91.9 44.8 99.7 89.1 89.3 90.6 97.4 95.9 83.9 97.6 91.8 84.4 59.0 92.5 74.3 97.4 86.4 99.7 86.9
L. Zhong et al. TIP (2020) 88.8 41.3 94.0 85.9 86.9 89.0 98.5 93.7 83.1 87.3 86.2 78.5 58.6 86.3 57.9 91.7 85.0 96.2 82.7
M. Stoiber et al. ACCV (2020) 96.4 53.2 98.8 93.9 93.0 92.7 99.7 97.1 92.5 92.5 93.7 88.5 70.0 92.1 78.8 95.5 92.5 99.6 90.0
X. Sun et al. TCSVT (2021) 67.2 25.2 80.7 27.6 80.2 56.6 70.1 52.9 17.3 85.3 37.8 53.0 36.3 57.9 24.6 68.0 82.6 86.7 56.1
X. Sun et al. (1) TCSVT (2021) 93.0 55.2 99.3 85.4 96.1 93.9 98.0 95.6 79.5 98.2 89.7 89.1 66.5 91.3 60.6 98.6 95.6 99.6 88.1
J. Li et al. C&G (2021) 90.0 47.4 88.4 82.1 82.9 77.9 94.2 89.3 79.2 71.0 86.3 79.3 48.5 66.5 55.9 77.1 77.6 92.0 77.0
J. Li et al. JCST (2021) 92.8 42.6 96.8 87.5 90.7 86.2 99.0 96.9 86.8 94.6 90.4 87.0 57.6 88.7 59.9 96.5 90.6 99.5 85.8
H. Huang et al. TVCG (2022) 94.6 49.4 99.5 91.0 93.7 96.0 97.8 96.6 90.2 98.2 93.4 90.3 64.4 94.0 79.0 98.8 92.9 99.8 89.9
X. Tian et al. (2) ECCV (2022) 99.8 67.1 100.0 97.8 97.3 93.7 100.099.4 97.4 97.6 99.3 96.9 84.7 97.7 93.4 96.7 95.4 100.0 95.2
X. Tian et al. (f) ECCV (2022) 99.9 65.8 99.4 95.4 96.7 92.4 100.0 98.7 94.8 97.0 97.3 96.2 82.9 96.9 91.0 95.9 94.9 99.8 94.2
F. Liu et al. TIE (2022) 93.7 39.3 98.4 91.6 84.6 89.2 97.9 95.9 86.3 95.1 93.4 77.7 61.5 87.8 65.0 95.2 85.7 99.8 85.5
M. Stoiber et al. IJCV (2022) 98.8 65.1 99.6 96.0 98.0 96.5 100.0 98.4 94.1 96.9 98.0 95.3 79.3 96.0 90.3 97.4 96.2 99.8 94.2
Q. Wang et al. TIE (2023) - - - - - - - - - - - - - - - - - - 94.6
X. Song et al. (3) CGF (2023) - - - - - - - - - - - - - - - - - - 95.4
N. Lv et al. AEI (2024) 97.7 54.2 99.6 96.9 99.1 98.7 100.0 99.5 95.6 98.9 96.9 96.9 83.0 97.1 95.3 97.6 97.6 99.8 94.7
K. Li et al. JEI (2025) - - - - - - - - - - - - - - - - - - 94.7
J. Chen et al. (*)(4) TIM (2025) 98.2 64.9 99.4 96.8 97.8 96.6 99.7 97.9 94.2 97.1 96.6 95.5 83.7 95.1 92.3 97.0 96.2 99.8 94.4
J. Chen et al. (*) TIM (2025) 99.0 85.0 99.6 97.5 99.6 98.1 100 98.8 97.5 98.6 98.9 98.5 87.4 97.2 95.4 98.1 98.7 100 97.1
(*) With an offlineprocessing stage
  1. When combined with [H. Tjaden et al.], TPAMI (2019)
  2. When S = 1
  3. When roughness = 0
  4. When contour-based method only
  1. The results were taken from [H. Huang et al.], CGF (2020).
  2. The results were taken from [X. Sun et al.], TCSVT (2021).
  3. The results were taken from [H. Tjaden et al.], TPAMI (2019).
  4. The results were taken from [H. Huang et al.], CGF (2020).
  5. The results were taken from [H. Huang et al.], CGF (2020).
  6. The results were taken from [N. Lv et al.], AEI (2024).

Dynamic Light scene

Method Ape Soda Vise Soup Camera Can Cat Clown Cube Driller Duck Egg Glue Iron Candy Lamp Phone Squirrel Avg.
G. Wang et al. (a) VC (2015) 13.1 12.2 15.3 15.9 14.4 14.2 18.8 12.3 14.0 15.8 16.2 12.7 13.0 14.9 8.8 14.3 12.4 18.1 14.2
G. Wang et al. (b) VC (2015) 5.7 3.9 6.5 8.4 4.7 5.4 18.5 2.9 3.3 6.4 6.9 3.1 2.6 5.0 1.5 3.4 3.5 12.6 5.8
H. Tjaden et al. (c) ICCV (2017) 61.7 32.0 94.2 66.3 68.0 84.1 96.6 85.8 45.7 88.7 74.1 56.9 29.9 49.1 20.7 91.5 63.0 98.5 67.0
B. Wang et al. (d) MTAP (2019) 20.7 19.9 23.7 21.5 23.5 22.4 20.1 21.9 22.7 23.8 19.9 22.4 18.2 24.4 20.8 22.4 24.4 23.6 22.0
L. Zhong et al. (e) IJCV (2019) 81.8 39.7 91.5 85.1 82.6 87.1 98.1 90.7 79.7 87.4 81.6 73.1 51.7 75.9 53.4 88.8 78.6 95.6 79.0
H. Tjaden et al. TPAMI (2019) 84.9 42.0 99.0 81.3 84.3 88.9 95.6 92.5 77.5 94.6 86.4 77.3 52.9 77.9 47.9 96.9 81.7 99.3 81.2
H. Huang et al. CGF (2020) 91.8 42.3 98.9 89.9 91.3 87.8 97.6 94.5 84.5 98.1 91.9 86.7 66.2 90.9 73.2 97.1 89.2 99.6 87.3
L. Zhong et al. TIP (2020) 89.7 40.2 92.7 86.5 86.6 89.2 98.3 93.9 81.8 88.4 83.9 76.8 55.3 79.3 54.7 88.7 81.0 95.8 81.3
M. Stoiber et al. ACCV (2020) 96.5 54.6 99.1 93.9 93.1 94.7 99.5 97.0 93.0 93.4 93.3 92.6 74.9 91.0 79.2 95.6 89.8 99.5 90.6
X. Sun et al. TCSVT (2021) 71.2 26.6 81.1 33.3 84.0 57.3 76.5 65.3 21.5 78.3 52.0 62.1 46.8 62.5 29.5 67.6 82.5 90.6 60.5
X. Sun et al. (1) TCSVT (2021) 93.8 55.9 99.6 85.6 97.7 93.7 97.7 96.5 78.3 98.6 91.0 91.6 72.1 90.7 63.0 98.9 94.4 100.0 88.8
J. Li et al. C&G (2021) 89.1 44.5 92.1 86.7 87.7 81.6 95.7 92.5 81.0 75.5 87.7 82.8 61.1 62.4 61.9 76.0 73.8 94.4 79.3
J. Li et al. JCST (2021) 93.5 43.1 96.6 88.5 92.8 86.0 99.6 95.5 85.7 96.8 91.1 90.2 68.4 86.8 59.7 96.1 91.5 99.2 86.7
H. Huang et al. TVCG (2022) 94.3 48.3 99.5 90.1 94.6 96.1 97.9 97.3 90.9 99.1 92.9 91.5 72.6 94.7 80.0 98.3 95.2 99.8 90.7
X. Tian et al. (2) ECCV (2022) 100.0 64.5 99.8 97.9 97.9 94.0 100.0 99.5 97.0 98.8 99.3 97.6 87.5 97.4 92.4 97.1 96.4 100.0 95.4
X. Tian et al. (f) ECCV (2022) 99.6 63.9 99.7 95.3 97.1 93.0 100.0 98.7 93.8 98.5 96.9 97.1 86.6 95.1 91.2 96.2 95.6 99.9 94.3
F. Liu et al. TIE (2022) 93.5 38.2 98.4 88.8 87.0 88.5 98.1 94.4 85.1 95.1 92.7 76.1 58.1 79.6 62.1 93.2 84.7 99.6 84.1
M. Stoiber et al. IJCV (2022) 98.2 65.2 99.2 95.6 97.5 98.1 100.0 98.5 94.2 97.5 97.9 96.9 86.1 95.2 89.3 97.0 95.9 99.9 94.6
Q. Wang et al. TIE (2023) - - - - - - - - - - - - - - - - - - 95.2
X. Song et al. (3) CGF (2023) - - - - - - - - - - - - - - - - - - 94.9
N. Lv et al. AEI (2024) 98.5 49.7 99.8 97.1 98.6 98.8 100.0 99.2 97.1 98.9 97.9 98.4 87.8 97.9 94.4 97.9 97.8 100.0 95.0
K. Li et al. JEI (2025) - - - - - - - - - - - - - - - - - - 95.2
J. Chen et al. (*)(4) TIM (2025) 98.8 64.1 99.6 96.4 97.7 97.8 99.5 98.2 94.7 97.4 96.8 97.1 88.5 94.8 92.3 97.2 96.4 99.6 94.8
J. Chen et al. (*) TIM (2025) 99.0 80.5 99.8 97.2 99.7 98.4 99.7 99.3 97.1 99.1 98.2 99.0 89.0 97.1 96.2 98.9 98.8 100 97.0
(*) With an offlineprocessing stage
  1. When combined with [H. Tjaden et al.], TPAMI (2019)
  2. When S = 1
  3. When roughness = 0
  4. When contour-based method only
  1. The results were taken from [H. Huang et al.], CGF (2020).
  2. The results were taken from [X. Sun et al.], TCSVT (2021).
  3. The results were taken from [H. Tjaden et al.], TPAMI (2019).
  4. The results were taken from [H. Huang et al.], CGF (2020).
  5. The results were taken from [H. Huang et al.], CGF (2020).
  6. The results were taken from [N. Lv et al.], AEI (2024).

Noisy (+ Dyanmic Light) scene

Method Ape Soda Vise Soup Camera Can Cat Clown Cube Driller Duck Egg Glue Iron Candy Lamp Phone Squirrel Avg.
G. Wang et al. (a) VC (2015) 10.4 10.3 12.3 14.5 6.5 8.4 16.6 11.4 6.9 9.3 17.5 3.7 5.6 11.4 5.8 9.6 7.3 12.9 10.0
G. Wang et al. (b) VC (2015) 2.5 2.6 1.1 5.1 0.7 0.2 5.6 1.4 0.7 0.2 7.2 0.1 0.6 1.1 1.0 0.8 0.2 2.0 1.9
H. Tjaden et al. (c) ICCV (2017) 55.9 35.3 75.4 67.4 27.8 10.2 94.3 33.4 8.6 50.9 76.3 2.3 2.2 18.2 11.4 36.6 31.3 93.5 40.6
B. Wang et al. (d) MTAP (2019) 19.8 21.4 21.2 21.0 22.0 20.3 19.8 21.2 22.4 20.4 20.4 20.7 19.1 20.4 20.9 20.6 21.0 20.8 20.7
L. Zhong et al. (e) IJCV (2019) 80.5 35.0 80.9 85.5 58.4 53.5 96.7 65.9 38.2 71.8 85.8 29.7 17.0 59.3 34.8 61.1 60.8 93.6 61.6
H. Tjaden et al. TPAMI (2019) 77.5 44.5 91.5 82.9 51.7 38.4 95.1 69.2 24.4 64.3 88.5 11.2 2.9 46.7 32.7 57.3 44.1 96.6 56.6
H. Huang et al. CGF (2020) 89.0 45.0 89.5 90.2 68.9 38.3 95.9 72.8 20.1 85.5 92.2 26.8 15.8 66.2 52.2 58.3 65.1 98.4 65.0
L. Zhong et al. TIP (2020) 79.3 35.2 82.6 86.2 65.1 56.9 96.9 67.0 37.5 75.2 85.4 35.2 18.9 63.7 35.4 64.6 66.3 93.2 63.6
M. Stoiber et al. ACCV (2020) 91.9 53.3 90.2 92.6 67.9 59.3 98.4 80.6 43.5 78.1 92.5 44.0 31.3 72.3 62.0 59.9 71.7 98.3 71.5
X. Sun et al. TCSVT (2021) 72.0 25.1 79.1 28.6 82.9 56.2 75.6 61.6 20.4 84.2 47.9 60.6 44.7 60.5 28.4 69.1 82.4 87.5 59.3
X. Sun et al. (1) TCSVT (2021) 92.5 56.2 98.0 85.1 91.7 79.0 97.7 86.2 40.1 96.6 90.8 70.2 50.9 84.3 49.9 91.2 89.4 99.4 80.5
J. Li et al. C&G (2021) 83.8 45.9 79.2 82.5 61.1 43.2 92.5 68.2 26.4 58.7 86.1 33.1 19.4 44.5 44.4 53.2 51.6 91.0 59.2
J. Li et al. JCST (2021) 89.1 44.0 91.6 89.4 75.2 62.3 98.6 77.3 41.2 81.5 91.6 54.5 31.8 65.0 46.0 78.5 69.6 97.6 71.4
H. Huang et al. TVCG (2022) 91.0 49.1 95.6 91.0 76.3 54.1 97.1 73.7 27.3 92.8 95.3 30.2 7.8 73.9 56.8 71.4 70.8 98.7 69.6
X. Tian et al. (2) ECCV (2022) 99.3 62.0 95.8 97.7 90.4 68.6 99.9 91.3 54.2 95.4 99.0 64.8 51.6 89.2 75.2 74.7 87.6 100.0 83.2
X. Tian et al. (f) ECCV (2022) 98.5 61.4 95.1 94.2 90.7 69.5 99.9 89.1 56.3 95.6 97.3 68.3 52.5 88.1 76.5 76.6 88.9 99.2 83.2
F. Liu et al. TIE (2022) 84.7 33.0 88.8 89.5 56.4 50.1 94.1 66.5 32.3 79.6 94.2 29.6 19.9 63.4 40.3 61.6 62.4 96.9 63.5
M. Stoiber et al. IJCV (2022) 96.9 61.9 95.4 95.7 84.5 73.9 99.9 90.3 62.2 87.8 97.6 62.2 43.4 84.3 78.2 73.3 83.1 99.7 81.7
Q. Wang et al. TIE (2023) - - - - - - - - - - - - - - - - - - 83.1
X. Song et al. (3) CGF (2023) - - - - - - - - - - - - - - - - - - 86.2
N. Lv et al. AEI (2024) 98.0 50.3 95.1 98.5 93.8 83.2 99.9 89.9 65.2 94.7 97.2 79.5 59.2 89.6 86.3 77.7 90.0 99.9 86.0
K. Li et al. JEI (2025) - - - - - - - - - - - - - - - - - - 83.4
J. Chen et al. (*)(4) TIM (2025) 98.4 60.9 95.9 95.4 91.3 80.0 98.9 87.7 65.3 94.8 96.6 71.4 54.8 88.3 79.4 83.2 89.4 99.6 85.1
J. Chen et al. (*) TIM (2025) 98.7 78.1 97.5 97.1 96.7 90.0 99.0 95.1 82.2 98.3 97.7 86.1 66.5 92.7 91.6 93.4 95.1 100.0 92.0
(*) With an offlineprocessing stage
  1. When combined with [H. Tjaden et al.], TPAMI (2019)
  2. When S = 1
  3. When roughness = 0
  4. When contour-based method only
  1. The results were taken from [H. Huang et al.], CGF (2020).
  2. The results were taken from [X. Sun et al.], TCSVT (2021).
  3. The results were taken from [H. Tjaden et al.], TPAMI (2019).
  4. The results were taken from [H. Huang et al.], CGF (2020).
  5. The results were taken from [H. Huang et al.], CGF (2020).
  6. The results were taken from [N. Lv et al.], AEI (2024).

Occlusion (+ Dynamic Light) scene

Method Ape Soda Vise Soup Camera Can Cat Clown Cube Driller Duck Egg Glue Iron Candy Lamp Phone Squirrel Avg.
G. Wang et al. (a) VC (2015) 12.1 12.0 14.6 13.9 13.5 14.7 18.4 13.1 13.2 16.8 17.0 11.6 11.6 15.1 8.0 14.3 14.3 16.7 13.9
G. Wang et al. (b) VC (2015) 5.8 3.2 6.7 8.2 6.3 5.1 13.9 3.6 3.6 5.6 8.0 2.6 2.8 4.6 1.5 3.6 4.5 9.8 5.5
H. Tjaden et al. (c) ICCV (2017) 55.2 29.9 82.4 56.9 55.7 72.2 87.9 75.7 39.6 78.7 68.1 47.1 26.2 35.6 16.6 78.6 50.3 77.6 57.5
B. Wang et al. (d) MTAP (2019) 19.8 21.4 21.2 21.0 22.0 20.3 19.8 21.2 22.4 20.4 20.4 20.7 19.1 20.4 20.9 20.6 21.0 20.8 20.7
L. Zhong et al. (e) IJCV (2019) 77.7 37.3 87.1 78.7 74.6 81.0 93.8 84.3 73.2 83.7 77.0 66.4 48.6 70.8 49.6 85.0 73.8 90.6 74.1
H. Tjaden et al. TPAMI (2019) 80.0 42.7 91.8 73.5 76.1 81.7 89.8 82.6 68.7 86.7 80.5 67.0 46.6 64.0 43.6 88.8 68.6 86.2 73.3
H. Huang et al. CGF (2020) 86.2 46.3 97.8 87.5 86.5 86.3 95.7 90.7 78.8 96.5 86.0 80.6 59.9 86.8 69.6 93.3 81.8 95.8 83.6
L. Zhong et al. TIP (2020) 83.9 38.1 92.4 81.5 81.3 85.5 97.5 88.9 76.1 87.5 81.7 72.7 52.5 77.2 53.9 88.5 79.3 92.5 78.4
M. Stoiber et al. ACCV (2020) 90.8 51.7 95.9 88.5 88.0 90.5 96.9 91.6 87.1 90.3 86.4 85.6 65.8 87.0 72.7 91.2 84.0 97.0 85.6
X. Sun et al. TCSVT (2021) 71.2 25.3 81.5 32.8 81.1 57.9 76.6 65.4 22.3 81.3 45.2 61.7 46.2 60.2 31.0 69.4 80.0 86.8 59.8
X. Sun et al. (1) TCSVT (2021) 91.3 56.7 97.8 82.0 92.8 89.9 96.6 92.2 71.8 97.0 85.0 84.6 66.9 87.7 56.1 95.1 89.8 98.2 85.1
J. Li et al. C&G (2021) 84.0 43.1 83.5 76.7 77.9 71.4 89.8 82.0 69.0 67.5 80.1 74.4 50.7 53.5 53.0 69.5 67.6 87.3 71.2
J. Li et al. JCST (2021) 89.3 43.3 92.2 83.1 84.1 79.0 94.5 88.6 76.2 90.4 87.0 80.7 61.6 75.3 53.1 91.1 81.9 93.4 80.3
H. Huang et al. TVCG (2022) 92.5 51.5 99.2 90.7 92.1 92.2 97.7 94.2 89.8 98.4 91.3 90.7 66.3 91.7 75.3 95.9 92.1 99.0 88.9
X. Tian et al. (2) ECCV (2022) 98.7 68.4 99.9 97.5 98.3 93.0 99.9 99.4 95.1 97.9 99.1 96.9 85.5 97.0 90.3 96.3 95.1 100.0 94.9
X. Tian et al. (f) ECCV (2022) 98.1 65.0 98.8 95.2 97.0 91.9 99.8 98.7 92.5 97.3 97.1 95.9 83.8 94.8 89.3 95.9 94.2 99.6 93.6
F. Liu et al. TIE (2022) 87.1 36.7 91.7 78.8 79.2 82.5 92.8 86.1 78.0 90.2 83.4 72.0 52.3 72.8 55.9 86.9 77.8 93.0 77.6
M. Stoiber et al. IJCV (2022) 96.5 66.8 99.0 95.8 95.0 95.9 100.0 97.6 92.2 96.6 95.0 94.4 79.0 94.7 89.8 95.7 93.6 99.6 93.2
Q. Wang et al. TIE (2023) - - - - - - - - - - - - - - - - - - 93.6
X. Song et al. (3) CGF (2023) - - - - - - - - - - - - - - - - - - 93.2
N. Lv et al. AEI (2024) 96.7 53.8 99.4 96.0 98.1 98.0 99.7 98.8 95.7 98.0 95.0 97.8 84.8 97.1 92.6 96.0 96.3 99.9 94.1
K. Li et al. JEI (2025) - - - - - - - - - - - - - - - - - - 94.5
J. Chen et al. (*)(4) TIM (2025) 96.2 62.6 98.5 94.4 96.1 96.2 99.1 96.9 92.3 97.0 94.0 95.0 84.1 93.7 89.9 95.6 94.4 98.9 93.1
J. Chen et al. (*) TIM (2025) 97.3 78.1 99.4 96.4 99.2 96.6 99.4 97.8 95.7 98.4 96.1 98.1 86.8 96.6 93.2 96.6 96.8 99.4 95.6
(*) With an offlineprocessing stage
  1. When combined with [H. Tjaden et al.], TPAMI (2019)
  2. When S = 1
  3. When roughness = 0
  4. When contour-based method only
  1. The results were taken from [H. Huang et al.], CGF (2020).
  2. The results were taken from [X. Sun et al.], TCSVT (2021).
  3. The results were taken from [H. Tjaden et al.], TPAMI (2019).
  4. The results were taken from [H. Huang et al.], CGF (2020).
  5. The results were taken from [H. Huang et al.], CGF (2020).
  6. The results were taken from [N. Lv et al.], AEI (2024).