Recurrent convolutional neural network for object recognition

By Ming Liang, Xiaolin Hu. 2015

1,016 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Electrical and Electronic Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Electrical and Electronic Engineering

Citations in the corpus

Deep Networks with Stochastic Depth (2016, neutral): “pConnect [20] 9.32 - 1.94 - Net in Net [24] 8.81 - 2.35 - Deeply Supervised [13] 7.97 - 1.92 33.70 Frac. Pool [25] - 27.62 - - All-CNN [6] 7.25 - - 41.20 Learning Activation [26] 7.51 30.83 - - R-CNN [27] 7.09 - 1.77 - Scalable BO [28] 6.37 27.40 1.77 - Highway Network [29] 7.60 32.24 - - Gen. Pool ”
Inner Monologue: Embodied Reasoning through Planning with Language Models (2022, background): “One common type of such feedback is object recognition [89, 90, 91, 92] that returns a list of”
FractalNet: Ultra-Deep Neural Networks without Residuals (2016, neutral): “ax-pooling and subsampling applied 5 Method C100 C100+ C100++ C10 C10+ C10++ SVHN Network in Network [21] 35.68 - - 10.41 8.81 - 2.35 Generalized Pooling [17] 32.37 - - 7.62 6.05 - 1.69 Recurrent CNN [19] 31.75 - - 8.69 7.09 - 1.77 Competitive Multi-scale [20] 27.56 - - 6.87 - - 1.76 FitNet [27] - 3”
Designing Neural Network Architectures using Reinforcement Learning (2016, neutral): “We posit that the CIFAR-10 MetaQNN Method CIFAR-10 SVHN MNIST CIFAR-100 DropConnect (Wan et al., 2013) 9.32 1.94 0.57 - DSN (Lee et al., 2015) 8.22 1.92 0.39 34.57 R-CNN (Liang & Hu, 2015) 7 (He et al., 2015) 6.61 - - - ELU (Clevert et al., 2015) 6.55 - - 24.28 Tree+Max-Avg (Lee et al., 2016) 6.05 1”
Understanding and Improving Convolutional Neural Networks via Concatenated Rectified Linear Units (2016, neutral): “We report the test set error rates Model CIFAR-10 CIFAR-100 (Rippel et al., 2015) 8.60 31.60 (Snoek et al., 2015) 6.37 27.40 (Liang & Hu, 2015) 7.09 31.75 (Lee et al., 2016) 6.05 32.37 (Srivastava et al., 2015) 7”
Large-Margin Softmax Loss for Convolutional Neural Networks (2016, background): “Error Rate FitNet (Romero et al., 2015) 35.04 NiN (Lin et al., 2014) 35.68 Maxout (Goodfellow et al., 2013) 38.57 DSN (Lee et al., 2015) 34.57 dasNet (Stollenga et al., 2014) 33.78 All-CNN (Springenberg et al., 2015) 33.71 R-CNN (Liang & Hu, 2015) 31.75 GenPool (Lee et al., 2016) 32.37 accuracy over”
Deeply-Recursive Convolutional Network for Image Super-Resolution (2015, background): “To overcome overﬁtting, Liang and Hu [17] uses a recurrent layer that takes feed-forward inputs into all un-folded layers.”
Batch-normalized Maxout Network in Network (2015, background): “Maxout network ( k =2) [7] 2.47 NIN [18] 2.35 Human performance [23] 2.00 MIM ( k =2) [17] 1.97 ± 0.08 Dropconnect [27] 1.94 DSN [15] 1.92 MIN ( k =5) 1.81 RCNN-192 [16] 1.77 with those obtained in recent studies.”
Beyond Shared Hierarchies: Deep Multitask Learning through Soft Layer Ordering (2017, neutral): “Since the type of recurrence induced by soft ordering does not require task input or output to be sequential, methods that use recurrence in such a setting are of particular interest (Liang and Hu, 2015; Liao and Poggio, 2016; Pinheiro and Collobert, 2014; Socher et al., 2011; Zamir et al., 2016).”
Noisy Softmax: Improving the Generalization Ability of DCNN via Postponing the Early Softmax Saturation (2017, methodology): “We use the average representation of 100 randomly Method MNIST CNN [19] 0.53 NiN [29] 0.47 Maxout [7] 0.45 DSN [27] 0.39 R-CNN [28] 0.31 GenPool [26] 0.31 DisturbLabel [51] 0.33 Softmax 0.43 Noisy Softmax ( α 2 = 1 ) 0.42 Noisy Softmax ( α 2 = 0 .”
Recent advances and clinical applications of deep learning in medical image analysis (2021, background): “Mask RCNN uses a Feature Pyramid Network (FPN) (Setio et al., 2017) as its backbone to extract features at various resolution scales.”
U-Net and Its Variants for Medical Image Segmentation: A Review of Theory and Applications (2020, methodology): “The recurrent U-net makes use of recurrent convolutional neural networks (RCNN) [63] by incorporating the recurrent feedback loops into a convolutional layer.”
Traffic Signs in the Wild: Highlights from the IEEE Video and Image Processing Cup 2017 Student Competition [SP Competitions] (2018, background): “Affiliation: Bangladesh University of Engineering and Technology Undergraduate students: Uday Kamal, Sowmitra Das, Abid Abrar Supervisor: Md Kamrul Hassan Technical Approach: Team Neurons developed a data-driven system based on Convolutional Neural Networks (CNNs) similar to an existing approach [5]”
Recent Advances in Recurrent Neural Networks (2017, background): “While applications of the MDRNN fall in line with RCNNs, there has yet to be any comparative examinations performed on the two models.”