Have anyone tried the latest released capsule network? Here is the paper: https://arxiv.org/pdf/1710.09829.pdf
\n\nBased on the method stats in the paper, it looks promising. There are many implementations online, but doesn't support different datasets. And don't have time create my own.
\n\nJust wonder how's the result would be on this dataset? Any specific reason using or not using it?
", "messages": [ { "id": 299302, "postDate": "2018-03-20T19:14:36.193Z", "content": "I used in Toxic comments challenge. They perform quite well. I guee they also would for speech recognition
", "rawMarkdown": "I used in Toxic comments challenge. They perform quite well. I guee they also would for speech recognition", "votes": 1, "replies": [ { "id": 299480, "postDate": "2018-03-20T22:19:08.443Z", "content": "I want to know your implementation after finishing the toxic comments challenge.\nI’m very interested in capsule network's usefulness in practice.
", "rawMarkdown": "I want to know your implementation after finishing the toxic comments challenge.\nI’m very interested in capsule network's usefulness in practice." } ] }, { "id": 295849, "postDate": "2018-03-14T09:56:34.353Z", "content": "yes.i am using capsule networks for traffic sign detection using a german traffic sign dataset
", "rawMarkdown": "yes.i am using capsule networks for traffic sign detection using a german traffic sign dataset", "votes": 1 }, { "id": 339266, "postDate": "2018-06-06T16:18:38.360Z", "content": "Hi, I'm using CapsuleNet (thanks to Xifeng Guo for his Keras implementation https://github.com/XifengGuo/CapsNet-Keras) and I've obtained up to 91 % of accuracy on the testset of the Speech Commands Data Set v0.01. Soon I'll try to evaluate the model on the challenge testset in order to compare it with the other results.
", "rawMarkdown": "Hi, I'm using CapsuleNet (thanks to Xifeng Guo for his Keras implementation https://github.com/XifengGuo/CapsNet-Keras) and I've obtained up to 91 % of accuracy on the testset of the Speech Commands Data Set v0.01. Soon I'll try to evaluate the model on the challenge testset in order to compare it with the other results.", "replies": [ { "id": 404280, "postDate": "2018-10-15T14:07:04.323Z", "rawMarkdown": "", "isDeleted": true }, { "id": 767273, "postDate": "2020-03-09T12:07:25.443Z", "content": "Iam using capsulenet too but my accuracy is not good. I got accuracy of 99% on testing and 85% in validation.iam experiencing overfitting. how can i solve the problem so that i get a perfect fit/
", "rawMarkdown": "Iam using capsulenet too but my accuracy is not good. I got accuracy of 99% on testing and 85% in validation.iam experiencing overfitting. how can i solve the problem so that i get a perfect fit/" } ] }, { "id": 336354, "postDate": "2018-05-31T13:55:25.060Z", "content": "i am also using capsule networks for traffic sign detection too..
", "rawMarkdown": "i am also using capsule networks for traffic sign detection too..", "replies": [ { "id": 338715, "postDate": "2018-06-05T16:24:03.513Z", "content": "you can refer https://github.com/dineshresearch/Novel-Deep-Learning-Model-for-Traffic-Sign-Detection-Using-Capsule-Networks
", "rawMarkdown": "you can refer https://github.com/dineshresearch/Novel-Deep-Learning-Model-for-Traffic-Sign-Detection-Using-Capsule-Networks\n", "votes": 1 } ] }, { "id": 269006, "postDate": "2018-01-16T03:41:09.627Z", "content": "Have anyone tried the latest released capsule network? Here is the paper: https://arxiv.org/pdf/1710.09829.pdf
\n\nBased on the method stats in the paper, it looks promising. There are many implementations online, but doesn't support different datasets. And don't have time create my own.
\n\nJust wonder how's the result would be on this dataset? Any specific reason using or not using it?
", "rawMarkdown": "Have anyone tried the latest released capsule network? Here is the paper: https://arxiv.org/pdf/1710.09829.pdf\n\nBased on the method stats in the paper, it looks promising. There are many implementations online, but doesn't support different datasets. And don't have time create my own.\n\nJust wonder how's the result would be on this dataset? Any specific reason using or not using it?" } ], "comments": [ { "id": 299302, "author_name": "Dieter", "author_url": "", "post_date": "2018-03-20T19:14:36.193000", "content": "I used in Toxic comments challenge. They perform quite well. I guee they also would for speech recognition
", "votes": 1, "replies": [ { "id": 299480, "author_name": "yuusha_aaaaa", "author_url": "", "post_date": "2018-03-20T22:19:08.443000", "content": "I want to know your implementation after finishing the toxic comments challenge.\nI’m very interested in capsule network's usefulness in practice.
", "votes": 0, "replies": [] } ] }, { "id": 295849, "author_name": "Amara Dinesh Kumar", "author_url": "", "post_date": "2018-03-14T09:56:34.353000", "content": "yes.i am using capsule networks for traffic sign detection using a german traffic sign dataset
", "votes": 1, "replies": [] }, { "id": 339266, "author_name": "FabioVesperini", "author_url": "", "post_date": "2018-06-06T16:18:38.360000", "content": "Hi, I'm using CapsuleNet (thanks to Xifeng Guo for his Keras implementation https://github.com/XifengGuo/CapsNet-Keras) and I've obtained up to 91 % of accuracy on the testset of the Speech Commands Data Set v0.01. Soon I'll try to evaluate the model on the challenge testset in order to compare it with the other results.
", "votes": 0, "replies": [ { "id": 404280, "author_name": "", "author_url": "", "post_date": "2018-10-15T14:07:04.323000", "content": "", "votes": 0, "replies": [] }, { "id": 767273, "author_name": "Jane Jepkoech", "author_url": "", "post_date": "2020-03-09T12:07:25.443000", "content": "Iam using capsulenet too but my accuracy is not good. I got accuracy of 99% on testing and 85% in validation.iam experiencing overfitting. how can i solve the problem so that i get a perfect fit/
", "votes": 0, "replies": [] } ] }, { "id": 336354, "author_name": "Gufran Miirza", "author_url": "", "post_date": "2018-05-31T13:55:25.060000", "content": "i am also using capsule networks for traffic sign detection too..
", "votes": 0, "replies": [ { "id": 338715, "author_name": "Amara Dinesh Kumar", "author_url": "", "post_date": "2018-06-05T16:24:03.513000", "content": "you can refer https://github.com/dineshresearch/Novel-Deep-Learning-Model-for-Traffic-Sign-Detection-Using-Capsule-Networks
", "votes": 1, "replies": [] } ] } ], "raw_markdown_by_id": { "299302": "I used in Toxic comments challenge. They perform quite well. I guee they also would for speech recognition", "295849": "yes.i am using capsule networks for traffic sign detection using a german traffic sign dataset", "339266": "Hi, I'm using CapsuleNet (thanks to Xifeng Guo for his Keras implementation https://github.com/XifengGuo/CapsNet-Keras) and I've obtained up to 91 % of accuracy on the testset of the Speech Commands Data Set v0.01. Soon I'll try to evaluate the model on the challenge testset in order to compare it with the other results.", "336354": "i am also using capsule networks for traffic sign detection too..", "269006": "Have anyone tried the latest released capsule network? Here is the paper: https://arxiv.org/pdf/1710.09829.pdf\n\nBased on the method stats in the paper, it looks promising. There are many implementations online, but doesn't support different datasets. And don't have time create my own.\n\nJust wonder how's the result would be on this dataset? Any specific reason using or not using it?" } }