Action Detection on JHMDB (split-1) benchmark leaderboard · SOTA2 Research

Action Detection on JHMDB (split-1)

79.1Brush Hair APfull

Evaluation Results

Method	Links
full2016.04		79.1	33.4	53.9	60.3	99.3	18.4	26.2	42	92.8	98.1	29.6	24.6	13.7	92.9	42.3	67.2	57.6	66.5	27.9	58.9	35.8	53.3
H-FCN2016.04		76.4	49.7	80.3	43	92.5	24.2	57.7	70.5	78.7	77.2	31.7	35.7	27	88.8	76.9	29.8	68.6	72.8	31.5	44.4	26.2	56.4
spatial-CNN2016.04		67.1	34.4	37.2	36.3	93.8	7.3	14.4	29.6	80.2	93.9	17.4	10	8.8	71.2	45.8	17.7	11.6	38.5	20.4	40.5	19.4	37.9
motion-CNN2016.04		66.3	16	60	51.6	88.6	18.9	10.8	23.9	83.4	96.7	18.2	17.2	14	84.4	19.3	72.6	61.8	76.8	17.3	46.7	14.3	45.7
A-FCN2016.04		66.2	45.7	54.6	42.2	83.9	4.2	33.5	31.7	75	76.6	24.8	18.5	28.3	82.3	70.8	18.2	32.6	31.7	31.7	23.9	18.8	42.6
full2016.04		65.2	18.3	38.1	39	79.4	7.3	9.4	25.2	80.2	82.8	33.6	11.6	5.6	66.8	27	32.1	34.2	33.6	15.5	34	21.9	36.2
M-FCN2016.04		64.2	38.1	80.1	39	91.8	34.7	57.4	74.6	74.5	77.6	31.3	40.9	18.5	89.4	59	32.3	69.3	82.9	25.8	46.1	22.2	54.8
H-FCN2016.04		60.1	34.2	56.4	38.9	83.1	10.8	24.5	38.5	71.5	67.5	21.3	19.8	11.6	78	50.6	10.9	43	48.9	26.5	25.2	15.8	39.9
A-FCN2016.04		56.5	34.7	40.1	43.1	76.9	2.7	17.7	15.6	71.2	51.5	17.9	12.4	12.9	65.4	53.3	5.3	16.4	22.6	27.6	13.2	15.3	32.5
spatial-CNN2016.04		55.8	25.5	25.1	24	77.5	1.9	21.4	68.6	71	15.4	6.3	4.6	41.1	28	9.4	8.2	19.9	17.8	29.2	11.5	27	27
M-FCN2016.04		42.9	19	49.6	28.9	71.8	14	20.4	36.6	60.1	66	18	17.3	8.3	73.5	26	11.6	44.1	53.7	17.6	22.4	11.5	34
motion-CNN2016.04		32.3	5	35.6	30.1	58	7.8	2.6	16.4	55	72.3	8.5	6.1	3.9	47.8	7.3	24.9	26.3	36.3	4.5	22.1	7.6	24.3