|
|
4e47429d82
|
Another part of OpenCV C-functions replaced by CPP-functions
|
2019-04-05 21:21:13 +03:00 |
|
|
|
f19d1440cb
|
Moved all OpenCV functions to the image_opencv.cpp/h
|
2019-04-04 22:57:05 +03:00 |
|
|
|
6cbf08321a
|
get rid of learning_rate_max
|
2019-03-19 00:02:01 +00:00 |
|
|
|
d64693eb77
|
add SGDR policy
|
2019-03-18 23:26:04 +00:00 |
|
|
|
b3254ed523
|
Fixed many warnings
|
2019-03-18 19:25:48 +03:00 |
|
|
|
dd0fdfb684
|
Compile fix
|
2019-03-02 03:37:55 +03:00 |
|
|
|
75f2a3e7cf
|
Added object Detection & Tracking using conv-rnn layer on frames from video
|
2019-03-02 03:32:24 +03:00 |
|
|
|
e82505d20e
|
Fixed random values
|
2019-02-28 16:55:30 +03:00 |
|
|
|
b3579380dc
|
improve compatibility with c++ compilers, prepare for CMake
|
2019-02-15 17:27:12 +01:00 |
|
|
|
28106c0fd8
|
Optimized memory allocation for XNOR on CPU
|
2019-02-12 22:16:11 +03:00 |
|
|
|
00e992a600
|
Compile fix
|
2019-02-12 02:12:46 +03:00 |
|
|
|
5448e07445
|
Try to fuse conv_xnor+shortcut -> conv_xnor
|
2019-02-12 02:05:15 +03:00 |
|
|
|
6832290eee
|
Fixed set_batch_network(), when workspace larger for smaller batch
|
2019-02-08 00:49:51 +03:00 |
|
|
|
c50b0e0c8a
|
Minor Python and C API improvement
|
2019-02-06 14:38:12 +03:00 |
|
|
|
7e9416aa80
|
Making a pointer version of network_predict for python.
The python binding requires the network struct to be passed
as a pointer to it rather than a struct copy.
|
2019-02-05 11:35:45 -06:00 |
|
|
|
8726d7b0db
|
Optimizing network_predict_image to resize only if necessary.
This speeds up the processing considerably if the user is so
kind to resize the image prior to doing detection.
|
2019-02-05 11:30:38 -06:00 |
|
|
|
c71354ab2e
|
Added cudaGetLastError() for cudaHostAlloc() to reset last cuda error
|
2019-01-31 14:22:07 +03:00 |
|
|
|
090d934c0f
|
Minor speedup on CPU
|
2019-01-26 19:12:46 +03:00 |
|
|
|
da044776d1
|
Merge pull request #2282 from davidssmith/master
add LSTM layer
|
2019-01-24 20:19:57 +03:00 |
|
|
|
5e778cd91e
|
add LSTM layer
|
2019-01-23 22:02:09 -06:00 |
|
|
|
8960fbfb3f
|
gcc on OSX required explicit return value for empty (char *) in detection_to_json
|
2019-01-22 10:17:42 -08:00 |
|
|
|
81f7fc2c7b
|
Fixed network resize memory allocation
|
2019-01-21 15:32:57 +03:00 |
|
|
|
0e022d0912
|
Fixed timer
|
2019-01-18 21:29:06 +03:00 |
|
|
|
bf6b40f4e9
|
Another CUDA performance improvements
|
2019-01-18 16:29:54 +03:00 |
|
|
|
5c55137687
|
Minor fix
|
2019-01-12 02:56:30 +03:00 |
|
|
|
95904cf65a
|
Bad memory allocation
the variable [network->seen is declared uint64_t* ](https://github.com/AlexeyAB/darknet/blob/master/include/darknet.h#L519), but constructing function allocates only 4 bytes for (int) here
|
2019-01-11 14:56:03 +02:00 |
|
|
|
5ee166032a
|
Use detector test...-out result.json -dont_show < train.txt to save results to JSON-file
|
2019-01-09 17:49:37 +03:00 |
|
|
|
e6eba22adb
|
Graceful shutdown closes sockets
|
2019-01-09 03:20:29 +03:00 |
|
|
|
c31a245665
|
Experimental. Use -json_port 8070 to allow connection to the port:8070 to get json-stream
|
2019-01-09 00:11:59 +03:00 |
|
|
|
3ff5084590
|
Added include/darknet.h
|
2019-01-06 23:51:38 +03:00 |
|
|
|
c75fbb5f2e
|
Minor fix
|
2019-01-06 15:45:10 +03:00 |
|
|
|
dc827f4c1c
|
Minor fix for mAP calculation during training
|
2018-12-27 00:30:05 +03:00 |
|
|
|
dc7f8a32ae
|
mAP calculation during training, if is used flag -map
|
2018-12-09 18:18:47 +03:00 |
|
|
|
25f65f6878
|
Added fast_binarize_weights_gpu()
|
2018-11-05 22:38:35 +03:00 |
|
|
|
c0e2512af2
|
Activation improvement, more robust timer.
|
2018-09-27 23:10:54 +03:00 |
|
|
|
7dd97537fb
|
XNOR-net tiny-yolo_xnor.cfg ~2x faster than cuDNN on CUDA (nVidia GPU Maxwell)
|
2018-09-22 02:01:14 +03:00 |
|
|
|
03e95320a1
|
XNOR coalesced memory access, and avoid bank conflicts
|
2018-09-17 23:39:25 +03:00 |
|
|
|
a9fef1bd66
|
Bug fixes. Tested im2col_cpu_custom_transpose - bad way.
|
2018-08-11 00:26:53 +03:00 |
|
|
|
d6162af210
|
Optimized on CPU: gemm_bin, im2col, activation, transpose
|
2018-08-09 02:31:36 +03:00 |
|
|
|
a284a7da8d
|
Try to use avx_hs() - slow and requires alignment 4096 bits < (l.size*l.size*l.c)
May be faster only from 8192 bits and more.
|
2018-08-08 19:08:58 +03:00 |
|
|
|
0a326e7afe
|
XNOR-net on CPU AVX2
|
2018-08-08 02:45:47 +03:00 |
|
|
|
cfc5fedbb6
|
Just used spaces for indents instead of Tabs
|
2018-07-10 23:29:15 +03:00 |
|
|
|
83b2ae66e1
|
Expose YOLODLL_API layer* get_network_layer(network* net, int i);
|
2018-06-10 09:45:39 +08:00 |
|
|
|
1c05ebf522
|
Minor fix
|
2018-06-07 03:39:30 +03:00 |
|
|
|
6c54f5ffb2
|
Show inconsistent information if it is present in .cfg and .names files
|
2018-06-07 02:20:44 +03:00 |
|
|
|
1e512b2308
|
Letter box is disabled in the darknet.py by defualt
|
2018-06-04 22:11:05 +03:00 |
|
|
|
51293290fa
|
Fix error in allocation size
|
2018-05-31 13:11:19 +03:00 |
|
|
|
160eddddc4
|
Minor fix
|
2018-05-27 16:53:42 +03:00 |
|
|
|
eef9f8e5bb
|
Minor fixes
|
2018-05-23 18:35:08 +03:00 |
|
|
|
89c11f83ed
|
Fixed darknet.py - uses batch=1 by default
|
2018-05-16 15:05:46 +03:00 |
|
