xinli-zn / tadt Goto Github PK
View Code? Open in Web Editor NEWImplementation of the TADT tracker of paper 'Target-Aware Deep Tracking'
Implementation of the TADT tracker of paper 'Target-Aware Deep Tracking'
你好,请问要运行这个测试效果,需要配置什么环境,需要matconvnet环境吗?Linux下可以吗
你好,编译matconvnet后,运行代码,报错错误使用 vl_nnconv
The option name is not a string (argument number 5),请问怎么解决呢?
你好,请问要运行这个测试效果,需要配置什么环境,需要matconvnet环境吗?Linux下可以吗
未定义函数或变量 'vl_nnconv'。
出错 dagnn.Conv/forward (line 11)
outputs{1} = vl_nnconv(...
出错 dagnn.Layer/forwardAdvanced (line 85)
outputs = obj.forward(inputs, {net.params(par).value}) ;
出错 dagnn.DagNN/eval (line 91)
obj.layers(l).block.forwardAdvanced(obj.layers(l)) ;
出错 get_subwindow_feature (line 37)
net_feat.eval({'input',gpuArray(patch_sw)});
出错 tadt_tracking (line 63)
feat_groups = get_subwindow_feature(net_feat, img, sw_location, input_sz, feat_layer);
出错 Demo_TADT (line 24)
[result, fps]=tadt_tracking(seq_info.img_list, seq_info.gt(1,:), vgg16_model_path, display);
The error like this :
Struct contents reference from a non-struct array object.
Error in solver.adam (line 62)
state.m = opts.beta1 * state.m + (1 - opts.beta1) * grad ;
Error in cnn_train_dag_ridge>accumulateGradients (line 375)
params.solver(net.params(p).value, state.solverState{p}, ...
Error in cnn_train_dag_ridge>processEpoch (line 282)
state = accumulateGradients(net, state, params, batchSize, parserv) ;
Error in cnn_train_dag_ridge (line 114)
[net, state, ders_iter(:,epoch)] = processEpoch(net, input, state, params, 'train') ;
I find that the params.solver is @solver.adam but the state.solverState{p} is blank in the below code. The code of initialize the state is enabled when the params.solver is empty.
if isempty(params.solver)
if isempty(state.solverState{p})
state.solverState{p} = zeros(size(parDer), 'like', parDer);
end
state.solverState{p} = vl_taccum(...
params.solverOpts.momentum, state.solverState{p}, ...
- (1 / batchSize), parDer) ;
net.params(p).value = vl_taccum(...
(1 - thisLR * thisDecay / (1 - params.solverOpts.momentum)), ...
net.params(p).value, ...
thisLR, state.solverState{p}) ;
else
grad = (1 / batchSize) * parDer + thisDecay * net.params(p).value;
% call solver function to update weights
[net.params(p).value, state.solverState{p}] = ...
params.solver(net.params(p).value, state.solverState{p}, ...
grad, params.solverOpts, thisLR) ;
end
so it cause a error when calling params.solver().
how should I fix it ?
Thanks !
I am getting the following error. Kindly help me to solve this issue
Demo_TADT
Attempt to execute SCRIPT vl_nnconv as a function:
G:\Documents\TADT-master\matconvnet\matlab\vl_nnconv.m
Error in dagnn.Conv/forward (line 11)
outputs{1} = vl_nnconv(...
Error in dagnn.Layer/forwardAdvanced (line 85)
outputs = obj.forward(inputs, {net.params(par).value}) ;
Error in dagnn.DagNN/eval (line 91)
obj.layers(l).block.forwardAdvanced(obj.layers(l)) ;
Error in get_subwindow_feature (line 37)
net_feat.eval({'input',gpuArray(patch_sw)});
Error in tadt_tracking (line 63)
feat_groups = get_subwindow_feature(net_feat, img, sw_location, input_sz, feat_layer);
Error in Demo_TADT (line 24)
[result, fps]=tadt_tracking(seq_info.img_list, seq_info.gt(1,:), vgg16_model_path, display);
您好,非常感谢你的工作,我下载了你提供的TC128的结果文件,但是发现只有128个文件,TC128共有129个seq,我查看了下貌似缺少了Railwaystation_ce的结果文件,不知道是否是这样?谢谢
What a great job! Can u tell me how to train?
我想用于多目标跟踪?请问如何实现
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.