17b22d7ce7
Update Readme.md
2018-03-01 03:10:32 +03:00
490d02505b
Fixed VideoWriter
2018-02-26 21:48:09 +03:00
e96a454ca1
Update Readme.md
2018-02-26 01:13:55 +03:00
48d713fd49
Fine-tuning densenet201_yolo.cfg and resnet50_yolo.cfg
2018-02-26 01:09:14 +03:00
cad4d1618f
Added support for Tensor Cores CC >= 7.0 (V100). For FP16/32 (mixed precision) define CUDNN_HALF should be used.
2018-02-25 16:29:44 +03:00
85eafd3d59
Added partial.cmd
2018-02-25 15:23:11 +03:00
3ff4797b4c
Added gen_anchors.py
2018-02-25 14:48:08 +03:00
175e47317e
Update Readme.md
2018-02-24 02:35:38 +03:00
cd2bdec090
Updated to CUDA 9.1. And fixed no_gpu dependecies.
2018-02-23 15:05:31 +03:00
6332ea99ab
one more fix
2018-02-23 00:13:08 +03:00
b2b5756d86
Added __float2half_rn() and __half2float()
2018-02-22 23:52:43 +03:00
f558d5c39c
Fix
2018-02-22 23:16:36 +03:00
dda993f3dd
Use half_float16 instead of float32 if defined both CUDNN and CUDNN_HALF. Use Tensor Cores.
2018-02-22 22:54:40 +03:00
033e934ce8
If there is excessive GPU-RAM consumption by CUDNN then then do not use Workspace
2018-02-21 19:14:01 +03:00
e4ab47dfce
Optimized resizing of region_layer for random=1
2018-02-21 15:35:09 +03:00
4b0be8c701
Optimized resizing of network for random=1
2018-02-21 15:06:11 +03:00
068090c3a4
Minor fix
2018-02-21 00:14:46 +03:00
2972f8de8e
Fixed Tracking using SparsePyrLKOpticalFlow on CPU
2018-02-20 02:43:04 +03:00
5f74aeb224
Merge pull request #398 from HowardStark/master
...
CPU implementation of Tracker_optflow
2018-02-20 01:53:45 +03:00
d8b59fcafe
Changed definition location to adhere to compatability promises
2018-02-19 14:26:53 -08:00
892860e84e
Added CPU Optical Flow
2018-02-18 22:04:01 -08:00
4a7a076d09
Update Readme.md
2018-02-19 02:28:46 +03:00
8474f491f2
Minor fix
2018-02-18 19:45:38 +03:00
1106f5325b
Update Readme.md
2018-02-18 19:44:58 +03:00
82f630cac7
Added param -http_port 8090 to show MJPEG-stream in the WebBrowser (Chrome/Firefox)
2018-02-18 02:31:56 +03:00
baf3fcb491
Calculation TP, FP, FN, avg_IoU for default threshold during mAP calculation
2018-02-17 02:26:08 +03:00
65bff2683b
It takes into account the Difficult for calculating mAP for PascalVOC
2018-02-16 23:55:37 +03:00
a1af57d8d6
Added C implementation of calculation mAP (mean average precision) using Darknet
2018-02-15 15:43:25 +03:00
6af2cf692c
resnet152.cfg
2018-02-15 14:52:51 +03:00
2c29218e4e
Added compute_mAP.cmd for calculation mAP for Pascal VOC 2007 dataset.
...
Added reval_voc_py3.py and voc_eval_py3.py for Python3.
2018-02-14 00:25:11 +03:00
9c847647a1
Another compile fix error
2018-02-12 22:54:10 +03:00
8df8b36f31
Compile fix
2018-02-12 22:49:40 +03:00
14db8f3384
Added linear extrapolation of coordinates
2018-02-12 22:36:52 +03:00
a2071e1585
Parameter (max) in the cfg-file has an effect on the Yolo-training
2018-02-10 01:48:38 +03:00
cd8a3dcb4c
Compile fixes
2018-02-09 02:22:42 +03:00
be90b8e8cb
Optimal params for optical flow tracking. Some small box fixes.
2018-02-09 01:50:35 +03:00
51d99f5903
Fixed compile error
2018-02-08 01:24:38 +03:00
68ff1ad368
Update Readme.md - parameter for small objects
2018-02-08 01:13:08 +03:00
f03d486760
Makefile added hints to compile for Jetson Tx1 and Tx2
2018-02-08 01:09:30 +03:00
4373f897f1
Fixed densenet201_yolo.cfg - burn_in and poly policy, that changes learning rate
2018-02-08 01:07:19 +03:00
8b385e1afe
Fix for tracking
2018-02-08 01:05:04 +03:00
9ff40baa4f
Fixes for training Yolo on small objects
2018-02-08 01:04:40 +03:00
64aa0180bb
Merge pull request #355 from PTS93/patch-1
...
Fix autosave interval
2018-02-03 15:42:16 +03:00
f762e6adb5
Merge pull request #357 from rajendraarora16/new-changes-darknet
...
Fixes Readme.md for darknet_no_gpu.exe
2018-02-03 15:38:31 +03:00
5baf36188a
Merge pull request #369 from stereomatchingkiss/fix_bug_368
...
fix bug 368
2018-02-03 15:37:48 +03:00
db3a3c54c5
Some fixes
2018-02-03 15:35:13 +03:00
192135106f
fix bug 368
2018-02-03 18:30:42 +08:00
063a23f637
Tracking small fixes
2018-01-31 01:55:25 +03:00
21a2c68da2
Fixes Readme.md for darknet_no_gpu
2018-01-29 18:27:34 +05:30
07b39c5f57
Fix autosave interval
...
In the current code it simply saves every 100 iterations. I changed the readme to reflect that.
In detector.c atm:
´//if (i % 1000 == 0 || (i < 1000 && i % 100 == 0)) {
if (i % 100 == 0) {
...
´
2018-01-28 14:28:30 +01:00
