ynma-hanvo
ynma-hanvo
hi, stupid question, what is the purpose of reference images provided by the author in pkl format . because when testing the network, it is using query.jpg
also want to export it to onnx ,but got no idea. any help?
i think this is a colmap issue. you may ask it on colmap repo.
same for me. I used stereo-Inertial with my stereo cameras, parameter from kalibr result. also tracking reset every few seconds. for stereo is ok
> 重建的深度只在0.5~50.0m之间。 如果深度很大，算法可能无法三角化50m以上的场景。 does the algorithm works well if we limit range to 0 - 15m , i tested it but seems it is still estimating depth for the values...
nice chat thread; I am wondering how long does it take for 768 * 480 depth inference , and is the output depth value the pixel shift between the left...
how is it for baseline about 10cm? can you share you results
> this paper seems like recovering rgb image at low light condition from rgbd depth .
> @hnsywangxin It's a ok. hello, what purpose do your use the disparity for , to me the generated disparity seems not accurate at all/
> Hi, @tv12345 I can run PSMNet on my 8GB Jetson TX2. A pair of 1242x375 images took around 4.5 GB memory. could you share how long does it take...