TontonTremblay/hope_pose_all_objects

HOPE Dataset — doper2 Pose Estimation (All 28 Objects)

Trained on train_pbr synthetic BOP data (~30k frames/object), evaluated on val. Backbone: ResNet-34, 64 SimCC keypoints, 640px detection, 256px crop, 100 epochs, fp32.

---

Val Metrics Summary

Obj	Det rate	PnP rate	ADD AUC	ADDS AUC	MSSD AUC	MSPD AUC	AR MSSD @20%	AR MSPD @40px	ADD mean (mm)	ADDS mean (mm)	Diameter (mm)
01	0.82	0.84	0.152	0.289	0.115	0.574	0.500	0.711	82.44	66.85	107.9
02	0.87	1.00	0.190	0.345	0.143	0.631	0.622	0.778	80.32	55.03	153.9
03	0.91	1.00	0.116	0.188	0.090	0.633	0.333	0.756	90.64	69.80	115.0
04	1.00	1.00	0.189	0.298	0.106	0.708	0.400	1.000	19.43	9.61	89.6
05	1.00	1.00	0.156	0.295	0.102	0.803	0.467	1.000	17.43	8.68	98.0
06	0.88	0.88	0.139	0.450	0.082	0.670	0.773	0.955	26.55	11.73	208.7
07	0.75	0.88	0.128	0.260	0.084	0.568	0.400	0.771	37.58	20.48	89.6
08	0.88	1.00	0.019	0.129	0.001	0.478	0.280	0.560	104.19	78.26	115.6
09	0.60	0.60	0.354	0.638	0.278	0.819	1.000	1.000	14.03	7.45	206.0
10	0.77	0.83	0.110	0.361	0.054	0.636	0.655	0.828	53.44	40.18	89.7
11	0.82	0.95	0.088	0.218	0.060	0.545	0.421	0.711	56.21	34.56	153.9
12	0.66	0.72	0.225	0.444	0.166	0.452	0.611	0.611	72.15	40.99	207.1
13	0.80	0.83	0.162	0.280	0.132	0.630	0.480	0.800	105.24	86.15	153.4
14	0.60	0.60	0.544	0.762	0.424	0.810	1.000	1.000	9.39	4.87	204.5
15	0.74	1.00	0.078	0.268	0.040	0.459	0.429	0.571	72.14	52.48	75.7
16	0.92	0.96	0.137	0.300	0.102	0.681	0.583	0.875	65.42	50.82	161.5
17	0.72	0.92	0.320	0.576	0.245	0.447	0.609	0.609	60.07	20.01	205.7
18	0.87	0.87	0.333	0.604	0.236	0.788	0.923	1.000	10.17	4.95	122.8
19	0.95	1.00	0.169	0.370	0.102	0.636	0.600	0.750	141.81	129.62	89.2
20	0.90	0.93	0.189	0.384	0.121	0.563	0.536	0.821	64.87	55.39	89.9
21	0.87	0.89	0.155	0.352	0.091	0.655	0.490	0.898	24.07	12.70	89.2
22	1.00	1.00	0.191	0.459	0.119	0.764	0.800	1.000	20.13	9.41	152.4
23	0.62	0.75	0.316	0.483	0.233	0.639	0.700	0.800	35.45	20.77	151.3
24	1.00	1.00	0.074	0.322	0.000	0.645	0.600	1.000	22.35	11.69	151.3
25	0.97	0.97	0.081	0.218	0.012	0.560	0.412	0.882	50.63	31.71	252.8
26	0.86	1.00	0.030	0.145	0.014	0.469	0.314	0.571	54.77	32.80	107.1
27	0.77	0.94	0.125	0.280	0.059	0.470	0.364	0.667	103.80	89.31	76.1
28	1.00	1.00	0.150	0.472	0.082	0.722	0.650	1.000	14.65	6.36	82.9

---

Per-Object Visualizations

Object 01

Val 3x3	Train 3x3

Object 02

Val 3x3	Train 3x3

Object 03

Val 3x3	Train 3x3

Object 04

Val 3x3	Train 3x3

Object 05

Val 3x3	Train 3x3

Object 06

Val 3x3	Train 3x3

Object 07

Val 3x3	Train 3x3

Object 08

Val 3x3	Train 3x3

Object 09

Val 3x3	Train 3x3

Object 10

Val 3x3	Train 3x3

Object 11

Val 3x3	Train 3x3

Object 12

Val 3x3	Train 3x3

Object 13

Val 3x3	Train 3x3

Object 14

Val 3x3	Train 3x3

Object 15

Val 3x3	Train 3x3

Object 16

Val 3x3	Train 3x3

Object 17

Val 3x3	Train 3x3

Object 18

Val 3x3	Train 3x3

Object 19

Val 3x3	Train 3x3

Object 20

Val 3x3	Train 3x3

Object 21

Val 3x3	Train 3x3

Object 22

Val 3x3	Train 3x3

Object 23

Val 3x3	Train 3x3

Object 24

Val 3x3	Train 3x3

Object 25

Val 3x3	Train 3x3

Object 26

Val 3x3	Train 3x3

Object 27

Val 3x3	Train 3x3

Object 28

Val 3x3	Train 3x3