{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":652642893,"defaultBranch":"main","name":"DFT-Density-Field-Transformer-from-multiple-Single-View-Predictions","ownerLogin":"keonhee-han","currentUserCanPush":false,"isFork":true,"isEmpty":false,"createdAt":"2023-06-12T13:50:05.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/39676134?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1695039681.0","currentOid":""},"activityList":{"items":[{"before":"148dd11a3759c57f8f63b92b1dfb4fa1c95344e4","after":"a22cc2653bc84d1f69df88e4808f584c6e37620f","ref":"refs/heads/main","pushedAt":"2024-05-08T20:20:07.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"Update README.md","shortMessageHtmlLink":"Update README.md"}},{"before":"2bd9586e331c85e64e59c564ddc977facefcb83f","after":"cf2c9e99c2fe525794b49f3061bb3a4c49b87347","ref":"refs/heads/kh","pushedAt":"2024-01-24T18:23:33.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"correction","shortMessageHtmlLink":"correction"}},{"before":"b823a5dadfa3ee73f6f2ec30ce38a1d3739b0b94","after":"2bd9586e331c85e64e59c564ddc977facefcb83f","ref":"refs/heads/kh","pushedAt":"2024-01-24T18:19:05.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"rebuttal_dynamic_object_histogram_comparison","shortMessageHtmlLink":"rebuttal_dynamic_object_histogram_comparison"}},{"before":"53235781806f4f6bac2506deff5ac6ba0ef998b5","after":"b823a5dadfa3ee73f6f2ec30ce38a1d3739b0b94","ref":"refs/heads/kh","pushedAt":"2023-12-05T14:15:43.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"checkpoint for CVPR","shortMessageHtmlLink":"checkpoint for CVPR"}},{"before":"044ea583b2643bea8456f2d768592eb84694dc63","after":"53235781806f4f6bac2506deff5ac6ba0ef998b5","ref":"refs/heads/kh","pushedAt":"2023-11-10T15:30:27.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"backup","shortMessageHtmlLink":"backup"}},{"before":"ffc51034f5a41ef53206179d4a286494c9c43e0c","after":"044ea583b2643bea8456f2d768592eb84694dc63","ref":"refs/heads/kh","pushedAt":"2023-11-06T20:48:50.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"working model for MVBTS training","shortMessageHtmlLink":"working model for MVBTS training"}},{"before":"6ee0e629b3c33d987cf11e4d6c61c0e665b8e3b0","after":"ffc51034f5a41ef53206179d4a286494c9c43e0c","ref":"refs/heads/kh","pushedAt":"2023-11-02T13:07:20.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"draft for pre-IBRNet and -Neuray","shortMessageHtmlLink":"draft for pre-IBRNet and -Neuray"}},{"before":"20338829d9975c9e799cc8e3c6d9edf31b3b8c55","after":"6ee0e629b3c33d987cf11e4d6c61c0e665b8e3b0","ref":"refs/heads/kh","pushedAt":"2023-10-30T10:27:52.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"implemented small MLP: confidence + denisty field prediciton","shortMessageHtmlLink":"implemented small MLP: confidence + denisty field prediciton"}},{"before":"4ccc2e3d6e6ab7103ff93d9f5c15873226481892","after":"20338829d9975c9e799cc8e3c6d9edf31b3b8c55","ref":"refs/heads/kh","pushedAt":"2023-10-19T21:52:42.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"Knowledge Distillation invalid ray handling","shortMessageHtmlLink":"Knowledge Distillation invalid ray handling"}},{"before":"3a2f8193f4aca03defe2cc30eee9ce2d614856e7","after":"4ccc2e3d6e6ab7103ff93d9f5c15873226481892","ref":"refs/heads/kh","pushedAt":"2023-10-17T11:22:01.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"final brush up for letting it run","shortMessageHtmlLink":"final brush up for letting it run"}},{"before":"ed6ddf82c3430a1fc99e0ac3b25ceae45ff78766","after":"3a2f8193f4aca03defe2cc30eee9ce2d614856e7","ref":"refs/heads/kh","pushedAt":"2023-10-16T19:37:27.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"code brush for configv2","shortMessageHtmlLink":"code brush for configv2"}},{"before":"0470343069207c7af74cba8013009b395ccc214f","after":"ed6ddf82c3430a1fc99e0ac3b25ceae45ff78766","ref":"refs/heads/kh","pushedAt":"2023-10-16T19:33:00.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"code brush for config","shortMessageHtmlLink":"code brush for config"}},{"before":"9f5030d4e7bd193eb922dbc0eab7518d0f93eece","after":"0470343069207c7af74cba8013009b395ccc214f","ref":"refs/heads/kh","pushedAt":"2023-10-16T19:28:58.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"ablation studies setting updated accordingly.","shortMessageHtmlLink":"ablation studies setting updated accordingly."}},{"before":"4c783ff6c35013ac05ace1d8011e212b559602b0","after":"9f5030d4e7bd193eb922dbc0eab7518d0f93eece","ref":"refs/heads/kh","pushedAt":"2023-10-16T09:01:35.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"additional features  for yaml config+ code brushup","shortMessageHtmlLink":"additional features for yaml config+ code brushup"}},{"before":"233ad720e3e17bc5cacefb3f4cedc114b84da275","after":"4c783ff6c35013ac05ace1d8011e212b559602b0","ref":"refs/heads/kh","pushedAt":"2023-10-15T18:10:22.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"training model is saved in Tensorboard file with overview of model configuration automatically. Hyper-paramter settings in configuration is added for mono and stereo setup","shortMessageHtmlLink":"training model is saved in Tensorboard file with overview of model co…"}},{"before":"8da960c2e60b99b92d9f3c8ed474f68abac4b644","after":"233ad720e3e17bc5cacefb3f4cedc114b84da275","ref":"refs/heads/kh","pushedAt":"2023-10-15T12:10:46.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"final dversion or MVBTS","shortMessageHtmlLink":"final dversion or MVBTS"}},{"before":"19e5829b50314e15090292101893bd4674c001c9","after":"8da960c2e60b99b92d9f3c8ed474f68abac4b644","ref":"refs/heads/kh","pushedAt":"2023-10-05T12:18:35.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"knowledge distillation: freezing MVHead during training in trainer initialization","shortMessageHtmlLink":"knowledge distillation: freezing MVHead during training in trainer in…"}},{"before":"70c1064d2de971aff8c2e06037be4ec6da936eda","after":"19e5829b50314e15090292101893bd4674c001c9","ref":"refs/heads/kh","pushedAt":"2023-10-05T08:12:48.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"Revert \"major change: fixing impossible training scheme for NeuRay: its visibility feature map called sampled_bottleneck_feats has activated computational graph during backpropagation, making out of CUDA memory. For that, I disabled by detaching it before it feeds into forward function for MVBTS head.\"\n\nThis reverts commit 194e60731a75561f24bd99254c89b79832c40fa7.","shortMessageHtmlLink":"Revert \"major change: fixing impossible training scheme for NeuRay: i…"}},{"before":"194e60731a75561f24bd99254c89b79832c40fa7","after":"70c1064d2de971aff8c2e06037be4ec6da936eda","ref":"refs/heads/kh","pushedAt":"2023-10-04T22:02:10.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"code brush up","shortMessageHtmlLink":"code brush up"}},{"before":"8db40e8972ab645ff8796c7d24d98902c38ba2f4","after":"194e60731a75561f24bd99254c89b79832c40fa7","ref":"refs/heads/kh","pushedAt":"2023-10-03T15:16:59.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"major change: fixing impossible training scheme for NeuRay: its visibility feature map called sampled_bottleneck_feats has activated computational graph during backpropagation, making out of CUDA memory. For that, I disabled by detaching it before it feeds into forward function for MVBTS head.","shortMessageHtmlLink":"major change: fixing impossible training scheme for NeuRay: its visib…"}},{"before":"6ec81621a6969a4977edfe61fc96280461057800","after":"8db40e8972ab645ff8796c7d24d98902c38ba2f4","ref":"refs/heads/kh","pushedAt":"2023-10-02T13:43:23.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"code brush up + fixed critically unrecognized configuration of dropout indices + added extra feature: token's embedding sequence order","shortMessageHtmlLink":"code brush up + fixed critically unrecognized configuration of dropou…"}},{"before":"8f41cd021b4c9730ea87068f2c533bc92ba8a78a","after":"6ec81621a6969a4977edfe61fc96280461057800","ref":"refs/heads/kh","pushedAt":"2023-09-29T19:18:29.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"hotfix: IBRAtt set to true for learning DFT model","shortMessageHtmlLink":"hotfix: IBRAtt set to true for learning DFT model"}},{"before":"fef2ad89f92a18b377487fec7d80676c8e1fc9b0","after":"8f41cd021b4c9730ea87068f2c533bc92ba8a78a","ref":"refs/heads/kh","pushedAt":"2023-09-28T13:15:33.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"testing validation of transformer working properly in DFT model: num view = 1","shortMessageHtmlLink":"testing validation of transformer working properly in DFT model: num …"}},{"before":"73e15448fc02c5cfec0417defdf4f91be56c46a3","after":"fef2ad89f92a18b377487fec7d80676c8e1fc9b0","ref":"refs/heads/kh","pushedAt":"2023-09-27T14:04:41.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"Fixed code mistake: invalid_features should take only one feature vector without dropout as it is pre-padded features","shortMessageHtmlLink":"Fixed code mistake: invalid_features should take only one feature vec…"}},{"before":"3241ea68afa6118f3e5acc73caeb68b339045fd0","after":"73e15448fc02c5cfec0417defdf4f91be56c46a3","ref":"refs/heads/kh","pushedAt":"2023-09-27T13:35:06.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"multiviewhead dropout feature added + code_brushed","shortMessageHtmlLink":"multiviewhead dropout feature added + code_brushed"}},{"before":"3246d6dbbad722c2cc1db3cd460134df45496f16","after":"3241ea68afa6118f3e5acc73caeb68b339045fd0","ref":"refs/heads/kh","pushedAt":"2023-09-26T10:32:00.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"code brush v3","shortMessageHtmlLink":"code brush v3"}},{"before":"0bb1460177f59e6d33148ee4d3b39db8c2a737c9","after":"3246d6dbbad722c2cc1db3cd460134df45496f16","ref":"refs/heads/kh","pushedAt":"2023-09-26T09:43:31.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"code brush v2","shortMessageHtmlLink":"code brush v2"}},{"before":"e636f74aee735de8a9aea45cd496e099869d1dac","after":"0bb1460177f59e6d33148ee4d3b39db8c2a737c9","ref":"refs/heads/kh","pushedAt":"2023-09-26T09:42:19.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"code brush after dominik meeting","shortMessageHtmlLink":"code brush after dominik meeting"}},{"before":"fc130c77a5a58d5b775458f5909701eb10fe3e56","after":"e636f74aee735de8a9aea45cd496e099869d1dac","ref":"refs/heads/kh","pushedAt":"2023-09-25T22:15:04.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"crucial update: previous feature encoding was constrained to only one feature map, but this is not proper way as all available views should be encoded into the feature maps. For that, **kwargs are added with name of mode such that the output inference from Resnetfc model will produce single view BTS encoded feature map. Then, this feature map can be converted to the density fields and compute the loss with MVBTS model","shortMessageHtmlLink":"crucial update: previous feature encoding was constrained to only one…"}},{"before":"198c909c92fcd04f18580dcefd7364386e36050e","after":"fc130c77a5a58d5b775458f5909701eb10fe3e56","ref":"refs/heads/kh","pushedAt":"2023-09-25T09:52:33.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"keonhee-han","name":"Kvasir8","path":"/keonhee-han","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39676134?s=80&v=4"},"commit":{"message":"code_brushup","shortMessageHtmlLink":"code_brushup"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAERQLUfwA","startCursor":null,"endCursor":null}},"title":"Activity · keonhee-han/DFT-Density-Field-Transformer-from-multiple-Single-View-Predictions"}