Hello,

thank you very much for the repo and the weights.

I want to use your model to inference an image.

Like you said i have to change "thermal.data" but i dont' have the txt-Files:
train = data/thermal_train.txt
valid = data/thermal_validation.txt

Could i have them ? Or is there another solution ? It seems like it worked for the other users. Am i missing something ?

@enesozi Please help me run the docker

../../../darknet: error while loading shared libraries: libcuda.so.1: cannot open shared object file: No such file or directory

Can you give me your dataset

Hi,

I am trying to depoly my computer.

When I command "make", I encountered compile fail.

I search a lot of related the problem.

I couldn't find the solution of it.

Please help me ~

The compile results the following :

nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -gencode arch=compute_61,code=[sm_61,compute_61] -Iinclude/ -I3rdparty/stb/include -DGPU -I/usr/local/cuda/include/ -DCUDNN --compiler-options "-Wall -Wfatal-errors -Wno-unused-result -Wno-unknown-pragmas -fPIC -ffp-contract=fast -mavx -mavx2 -msse3 -msse4.1 -msse4.2 -msse4a -Ofast -fopenmp -DGPU -DCUDNN -I/usr/local/cudnn/include -fPIC" -c ./src/maxpool_layer_kernels.cu -o obj/maxpool_layer_kernels.o
nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -gencode arch=compute_61,code=[sm_61,compute_61] -Iinclude/ -I3rdparty/stb/include -DGPU -I/usr/local/cuda/include/ -DCUDNN --compiler-options "-Wall -Wfatal-errors -Wno-unused-result -Wno-unknown-pragmas -fPIC -ffp-contract=fast -mavx -mavx2 -msse3 -msse4.1 -msse4.2 -msse4a -Ofast -fopenmp -DGPU -DCUDNN -I/usr/local/cudnn/include -fPIC" -c ./src/network_kernels.cu -o obj/network_kernels.o
nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -gencode arch=compute_61,code=[sm_61,compute_61] -Iinclude/ -I3rdparty/stb/include -DGPU -I/usr/local/cuda/include/ -DCUDNN --compiler-options "-Wall -Wfatal-errors -Wno-unused-result -Wno-unknown-pragmas -fPIC -ffp-contract=fast -mavx -mavx2 -msse3 -msse4.1 -msse4.2 -msse4a -Ofast -fopenmp -DGPU -DCUDNN -I/usr/local/cudnn/include -fPIC" -c ./src/avgpool_layer_kernels.cu -o obj/avgpool_layer_kernels.o
g++ -Iinclude/ -I3rdparty/stb/include -DGPU -I/usr/local/cuda/include/ -DCUDNN -Wall -Wfatal-errors -Wno-unused-result -Wno-unknown-pragmas -fPIC -ffp-contract=fast -mavx -mavx2 -msse3 -msse4.1 -msse4.2 -msse4a -Ofast -fopenmp -DGPU -DCUDNN -I/usr/local/cudnn/include -fPIC obj/http_stream.o obj/gemm.o obj/utils.o obj/dark_cuda.o obj/convolutional_layer.o obj/list.o obj/image.o obj/activations.o obj/im2col.o obj/col2im.o obj/blas.o obj/crop_layer.o obj/dropout_layer.o obj/maxpool_layer.o obj/softmax_layer.o obj/data.o obj/matrix.o obj/network.o obj/connected_layer.o obj/cost_layer.o obj/parser.o obj/option_list.o obj/darknet.o obj/detection_layer.o obj/captcha.o obj/route_layer.o obj/writing.o obj/box.o obj/nightmare.o obj/normalization_layer.o obj/avgpool_layer.o obj/coco.o obj/dice.o obj/yolo.o obj/detector.o obj/layer.o obj/compare.o obj/classifier.o obj/local_layer.o obj/swag.o obj/shortcut_layer.o obj/activation_layer.o obj/rnn_layer.o obj/gru_layer.o obj/rnn.o obj/rnn_vid.o obj/crnn_layer.o obj/demo.o obj/tag.o obj/cifar.o obj/go.o obj/batchnorm_layer.o obj/art.o obj/region_layer.o obj/reorg_layer.o obj/reorg_old_layer.o obj/super.o obj/voxel.o obj/tree.o obj/yolo_layer.o obj/upsample_layer.o obj/lstm_layer.o obj/image_opencv.o obj/convolutional_kernels.o obj/activation_kernels.o obj/im2col_kernels.o obj/col2im_kernels.o obj/blas_kernels.o obj/crop_layer_kernels.o obj/dropout_layer_kernels.o obj/maxpool_layer_kernels.o obj/network_kernels.o obj/avgpool_layer_kernels.o -o darknet -lm -pthread -lgomp -L/usr/local/cuda/lib64 -lcuda -lcudart -lcublas -lcurand -L/usr/local/cudnn/lib64 -lcudnn -lstdc++
obj/network.o: In function resize_network': network.c:(.text+0x1b87): undefined reference to resize_conv_lstm_layer'
network.c:(.text+0x1d27): undefined reference to resize_gaussian_yolo_layer' network.c:(.text+0x1da7): undefined reference to resize_scale_channels_layer'
network.c:(.text+0x1dbb): undefined reference to resize_sam_layer' obj/network.o: In function num_detections':
network.c:(.text+0x2457): undefined reference to gaussian_yolo_num_detections' obj/network.o: In function fill_network_boxes':
network.c:(.text+0x2d93): undefined reference to get_gaussian_yolo_detections' obj/network.o: In function free_network_recurrent_state':
network.c:(.text+0x514e): undefined reference to free_state_conv_lstm' obj/network.o: In function randomize_network_recurrent_state':
network.c:(.text+0x51fe): undefined reference to randomize_state_conv_lstm' obj/network.o: In function remember_network_recurrent_state':
network.c:(.text+0x52ae): undefined reference to remember_state_conv_lstm' obj/network.o: In function restore_network_recurrent_state':
network.c:(.text+0x532e): undefined reference to restore_state_conv_lstm' obj/parser.o: In function parse_conv_lstm':
parser.c:(.text+0x12e2): undefined reference to make_conv_lstm_layer' obj/parser.o: In function parse_gaussian_yolo':
parser.c:(.text+0x2a9b): undefined reference to make_gaussian_yolo_layer' obj/parser.o: In function parse_scale_channels':
parser.c:(.text+0x4f14): undefined reference to make_scale_channels_layer' obj/parser.o: In function parse_sam':
parser.c:(.text+0x503f): undefined reference to `make_sam_layer'
collect2: error: ld returned 1 exit status
Makefile:144: recipe for target 'darknet' failed
make: *** [darknet] Error 1

I ran ./build_docker_container.sh on my Jetson Xavier and received the error:

Step 5/9 : RUN apt-get update && apt-get install -y --no-install-recommends build-essential cmake git wget sudo python libcudnn7=$CUDNN_VERSION-1+cuda10.0 libcudnn7-dev=$CUDNN_VERSION-1+cuda10.0 && apt-mark hold libcudnn7 && rm -rf /var/lib/apt/lists/*
---> Running in 92c8d8acab95
standard_init_linux.go:211: exec user process caused "exec format error"
The command '/bin/sh -c apt-get update && apt-get install -y --no-install-recommends build-essential cmake git wget sudo python libcudnn7=$CUDNN_VERSION-1+cuda10.0 libcudnn7-dev=$CUDNN_VERSION-1+cuda10.0 && apt-mark hold libcudnn7 && rm -rf /var/lib/apt/lists/*' returned a non-zero code: 1

Hi @enesozi !
I am trying to recreate the thermal inference just like your video
https://youtu.be/xe_KRQ7qtQE

If I just want to do the inference.
Is it correct that I just need to change to use

1. thermal.data
2. thermal.names
3. yolov3-thermal.cfg
4. yolov3-spp-thermal .weights ?

Also, may I ask which python file do you use to create the inference result?
Since I should pass a single channel image to darknet. I am not quite should how should I modify it.

Please help! Thanks in advance

Hello Enes,

Thank you for your great medium post ! Could you provide the final weights you obtained for the YOLO classifier ?

Thank you !

I would like to give it a run but I'm out of a powerful GPU atm.

$ sudo ./run_docker_container.sh
Unable to find image 'thermal:darknet' locally
docker: Error response from daemon: pull access denied for thermal, repository does not exist or may require 'docker login'.

Ideas? Thank you.

Hi @enesozi

You mentioned you sampled the large objects from the train set. Could you shed some light, how did you do that? I am training model on the same dataset using yolov3-spp and just training the final layers as I am short of resources to train from scratch. Getting not so good results.

Hi @enesozi,

I was wondering if you have tried running inference on the frames of FLIR thermal video. I tried doing that using my trained weights and also yours, and for some reason, none of the weights are able to detect anything.
Wanted to know if someone had a success with your trained weights. Thanks!