HRSC reimplementations for other methods

[xiaoyihit]

I wonder if there is an MMRotate reimplementation for HRSC with other methods aside from s2anet and rotated retinanet.
I tried to replace the code in line 54 loss_bbox=dict(type='L1Loss', loss_weight=1.0)), in
configs/rotated_retinanet/rotated_retinanet_obb_r50_fpn_3x_hrsc_le90.py with
loss_bbox=dict(type='GDLoss', loss_type='gwd', loss_weight=5.0)),
loss_bbox=dict(type='KFLoss', loss_weight=5.0)),

loss_bbox=dict(
    type='GDLoss_v1',
    loss_type='kld',
    fun='log1p',
    tau=1,
    loss_weight=1.0)),

respectively, however none of them worked...
Their loss goes nan.
Guess I could get some help here?

[yangxue0827]

Refer to #183 and #202

HRSC

Backbone	mAP	AP50	AP75	Angle	lr schd	Mem (GB)	Inf Time (fps)	Aug	Batch Size	Configs	Download
ResNet50 (800,512)	52.06	84.80	58.10	le90	6x	1.56	38.2	RR	2	rotated_retinanet_obb_r50_fpn_6x_hrsc_rr_le90	model | log
ResNet50 (800,512)	54.15	86.20	60.60	le90	6x	1.56	38.2	RR	2	rotated_retinanet_obb_kld_stable_r50_fpn_6x_hrsc_rr_le90	model | log
ResNet50 (800,512)	45.09	79.30	46.90	oc	6x	1.56	39.2	RR	2	rotated_retinanet_hbb_r50_fpn_6x_hrsc_rr_oc	model | log
ResNet50 (800,512)	58.17	87.00	69.30	oc	6x	1.56	39.5	RR	2	rotated_retinanet_hbb_kld_stable_r50_fpn_6x_hrsc_rr_oc	model | log

[yangxue0827]

gwd and kfiou have not been reproduced on hrsc yet, but I think it should be similar to kld. Someone in Issue #202 ran successfully ran gwd on hrsc.

[yangxue0827]

copy from #202