Search results

<Thumos experiment>the python/tuple in train.prototxt was [2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38,40,42,44,46,48,50,52,54,56],which is different from [2,4,5,6,8,9,10,12,14,16] in your paper of R-C3D

Status: Open.

#64 In VisionLearningGroup/R-C3D;

· fanxx17 opened on Jan 4, 2021

<Thumos experiment> the dropout layer in train.prototxt was cancelled ,but i found overfitting when trying to test my caffemodel by using eval_detection.py

Status: Open.

#63 In VisionLearningGroup/R-C3D;

· fanxx17 opened on Jan 4, 2021

I cannot find this file '/scratch/detection/scnn/experiments/my_experiments/annotation/annotation_test/detclasslist.txt' when I test my model

Status: Open.

#62 In VisionLearningGroup/R-C3D;

· fanxx17 opened on Jan 4, 2021

Make error while building R-C3D/lib folder. nvcc error

Status: Open.

#59 In VisionLearningGroup/R-C3D;

· jahab opened on Dec 6, 2019

Can anybody share a demo result txt file for evaluation?

Status: Open.

#58 In VisionLearningGroup/R-C3D;

· dongzhuoyao opened on May 30, 2019

about batch_size

Status: Open.

#56 In VisionLearningGroup/R-C3D;

· lrhjw opened on May 16, 2019

about trainning my own dataset

Status: Open.

#55 In VisionLearningGroup/R-C3D;

· lrhjw opened on Apr 3, 2019

could you share the method to evaluation on charades

Status: Open.

#54 In VisionLearningGroup/R-C3D;

· GuardSkill opened on Mar 23, 2019

When I trained my own video base on "activitynet_iter_30000_3fps.caffemodel",it stoped at some one iteration.

Status: Open.

#53 In VisionLearningGroup/R-C3D;

· yp2005 opened on Mar 6, 2019

will 1st-5th groups C3D weights modified via back-propagation?

Status: Open.

#52 In VisionLearningGroup/R-C3D;

· rambleramble opened on Jan 29, 2019

could you share the method to generate "thumos14_gt_test.json"?

Status: Open.

#51 In VisionLearningGroup/R-C3D;

· rambleramble opened on Jan 28, 2019

Could I share your trained well models for thumos'14 and ActivityNet v1.3 datasets in your R-C3D paper?

Status: Open.

#50 In VisionLearningGroup/R-C3D;

· junmuzi opened on Jan 18, 2019