| 2026 | ICPR | Knowledge Distillation Through Low-Frequency Logits. | Takumi Kobayashi |
| 2026 | ICPR | MOST: Momentum Online Subspace Training. | Lincon S. Souza, Bojan Batalo, Takumi Kobayashi |
| 2026 | ICPR | Physics-Guided Prune-then-Finetune of Vision Transformers for Wavefield Pattern Analysis. | Jiaxing Ye, Takumi Kobayashi |
| 2025 | ICASSP | Matrix Decomposition By Additive and Subtractive Factors. | Takumi Kobayashi, Kenji Watanabe |
| 2025 | ICCV | Temperature in Cosine-based Softmax Loss. | Takumi Kobayashi |
| 2025 | ICIP | Geometric Mean Improves Loss For Few-Shot Learning. | Tong Wu, Takumi Kobayashi |
| 2025 | RO-MAN | Bi-LAT: Bilateral Control-Based Imitation Learning via Natural Language and Action Chunking with Transformers. | Takumi Kobayashi, Masato Kobayashi, Thanpimon Buamanee, Yuki Uranishi |
| 2024 | BMVC | Direct-Sum Approach to Integrate Losses Via Classifier Subspace. | Takumi Kobayashi |
| 2024 | CVPR | Mean-Shift Feature Transformer. | Takumi Kobayashi |
| 2024 | ICASSP | Local Distance Correlation Embedding for Time-Series Analysis on Riemannian Manifolds. | Lincon S. Souza, Takumi Kobayashi, Yasunori Nishimori, Yasuko Sugase-Miyamoto, Kenji Kawano, Shotaro Akaho, Narihisa Matsumoto |
| 2024 | WACV | Spatio-temporal Filter Analysis Improves 3D-CNN For Action Classification. | Takumi Kobayashi, Jiaxing Ye |
| 2023 | BMVC | Domain-Sum Feature Transformation For Multi-Target Domain Adaptation. | Takumi Kobayashi, Lincon S. Souza, Kazuhiro Fukui |
| 2023 | CVPR | Two-Way Multi-Label Loss. | Takumi Kobayashi |
| 2023 | ICIP | End-to-End Trainable Weakly Non-Negative Factorization. | Takumi Kobayashi, Kenji Watanabe |
| 2022 | BMVC | Mutual Conditional Probability for Self-Supervised Learning. | Takumi Kobayashi |
| 2022 | ECCV | Rotation Regularization Without Rotation. | Takumi Kobayashi |
| 2022 | WACV | Extractive Knowledge Distillation. | Takumi Kobayashi |
| 2021 | CVPR | T-vMF Similarity for Regularizing Intra-Class Feature Distribution. | Takumi Kobayashi |
| 2021 | WACV | Group Softmax Loss with Discriminative Feature Grouping. | Takumi Kobayashi |
| 2021 | WACV | Phase-wise Parameter Aggregation For Improving SGD Optimization. | Takumi Kobayashi |
| 2020 | CVPR | An Interface between Grassmann manifolds and vector spaces. | Lincon Sales de Souza, Naoya Sogi, Bernardo B. Gatto, Takumi Kobayashi, Kazuhiro Fukui |
| 2020 | ICIP | SCW-SGD: Stochastically Confidence-Weighted SGD. | Takumi Kobayashi |
| 2019 | BMVC | Large Margin In Softmax Cross-Entropy Loss. | Takumi Kobayashi |
| 2019 | ICCV | Global Feature Guided Local Pooling. | Takumi Kobayashi |
| 2019 | ICONIP | Simple ConvNet Based on Bag of MLP-Based Local Descriptors. | Takumi Kobayashi, Hidenori Ide, Kenji Watanabe |
| 2018 | ACCV | Spiral-Net with F1-Based Optimization for Image-Based Crack Detection. | Takumi Kobayashi |
| 2018 | BMVC | GAN-based Semi-supervised Learning On Fewer Labeled Samples. | Takumi Kobayashi |
| 2018 | CVPR | Analyzing Filters Toward Efficient ConvNet. | Takumi Kobayashi |
| 2018 | ICASSP | Trainable Co-Occurrence Activation Unit for Improving Convnet. | Takumi Kobayashi |
| 2018 | WISA | Secure Comparison Protocol with Encrypted Output and the Computation for Proceeding 2 Bits-by-2 Bits. | Takumi Kobayashi, Keisuke Hakuta |
| 2017 | BMVC | Feature Sequence Representation Via Slow Feature Analysis For Action Classification. | Takumi Kobayashi |
| 2017 | ICCV | Flip-Invariant Motion Representation. | Takumi Kobayashi |
| 2017 | ICONIP | Sharing ConvNet Across Heterogeneous Tasks. | Takumi Kobayashi |
| 2016 | BMVC | Learning Additive Kernel For Feature Transformation and Its Application to CNN Features. | Takumi Kobayashi |
| 2016 | CVPR | Structured Feature Similarity with Explicit Feature Map. | Takumi Kobayashi |
| 2016 | ICASSP | Histogram feature deblurring. | Takumi Kobayashi |
| 2016 | ICASSP | Discriminatively learned filter bank for acoustic features. | Takumi Kobayashi, Jiaxing Ye |
| 2016 | ICPR | Learning data-driven image similarity measure. | Takumi Kobayashi |
| 2015 | CAIP | Discriminative Local Binary Pattern for Image Feature Extraction. | Takumi Kobayashi |
| 2015 | CVPR | Three viewpoints toward exemplar SVM. | Takumi Kobayashi |
| 2015 | IJCCI | Pedestrian Action Prediction using Static Image Feature. | Kenji Nishida, Takumi Kobayashi, Taro Iwamoto, Shinya Yamasaki |
| 2015 | IJCCI | Preindication Mining for Predicting Pedestrian Action Change. | Kenji Nishida, Takumi Kobayashi, Taro Iwamoto, Shinya Yamasaki |
| 2014 | CVPR | Dirichlet-Based Histogram Feature Transform for Image Classification. | Takumi Kobayashi |
| 2014 | ICASSP | Acoustic feature extraction by statistics based local binary pattern for environmental sound classification. | Takumi Kobayashi, Jiaxing Ye |
| 2014 | ICASSP | Robust acoustic feature extraction for sound classification based on noise reduction. | Jiaxing Ye, Takumi Kobayashi, Masahiro Murakawa, Tetsuya Higuchi |
| 2014 | ICPR | S3CCA: Smoothly Structured Sparse CCA for Partial Pattern Matching. | Takumi Kobayashi |
| 2014 | ICPRAM | Discriminative Prior Bias Learning for Pattern Classification. | Takumi Kobayashi, Kenji Nishida |
| 2014 | ICPRAM | Learning Prior Bias in Classifier. | Takumi Kobayashi, Kenji Nishida |
| 2014 | ICPRAM | Tracking by Shape with Deforming Prediction for Non-rigid Objects. | Kenji Nishida, Takumi Kobayashi, Jun Fujiki |
| 2013 | CVPR | BFO Meets HOG: Feature Extraction Based on Histograms of Oriented p.d.f. Gradients for Image Classification. | Takumi Kobayashi |
| 2013 | ICASSP | Lp-regularized optimization by using orthant-wise approach for inducing sparsity. | Takumi Kobayashi |
| 2013 | ICASSP | Kernel discriminant analysis for environmental sound recognition based on acoustic subspace. | Jiaxing Ye, Takumi Kobayashi, Masahiro Murakawa, Tetsuya Higuchi |
| 2013 | Interspeech | Incremental acoustic subspace learning for voice activity detection using harmonicity-based features. | Jiaxing Ye, Takumi Kobayashi, Masahiro Murakawa, Tetsuya Higuchi |
| 2012 | ACCV | Generalized Mutual Subspace Based Methods for Image Set Classification. | Takumi Kobayashi |
| 2012 | BMVC | Higher-order Co-occurrence Features based on Discriminative Co-clusters for Image Classification. | Takumi Kobayashi |
| 2012 | ECCV | Efficient Optimization for Low-Rank Integrated Bilinear Classifiers. | Takumi Kobayashi, Nobuyuki Otsu |
| 2012 | ECCV | Efficient Similarity Derived from Kernel-Based Transition Probability. | Takumi Kobayashi, Nobuyuki Otsu |
| 2011 | ICASSP | Rotation invariant feature extraction from 3-D acceleration signals. | Takumi Kobayashi, Kiti Hasida, Nobuyuki Otsu |
| 2011 | ICASSP | Detection of peptide ion peaks in mass spectra by using weighted auto-correlation. | Kenji Watanabe, Takumi Kobayashi, Katsuyuki Koike, Tetsuya Higuchi, Tohru Natsume, Nobuyuki Otsu |
| 2011 | ICIP | Cancer detection from biopsy images using probabilistic and discriminative features. | Atsushi Yaguchi, Takumi Kobayashi, Kenji Watanabe, Kenji Iwata, Tadaaki Hosaka, Nobuyuki Otsu |
| 2011 | ICMLA | L0-Regularized Parametric Non-negative Factorization for Analyzing Composite Signals. | Takumi Kobayashi, Kenji Watanabe, Nobuyuki Otsu |
| 2011 | ICMLA | Efficient Optimization of Logistic Regression by Direct Use of Conjugate Gradient. | Kenji Watanabe, Takumi Kobayashi, Nobuyuki Otsu |
| 2010 | ICIP | Cone-restricted kernel subspace methods. | Takumi Kobayashi, Fumito Yoshikawa, Nobuyuki Otsu |
| 2010 | ICIP | Pyramidal segmentation using higher-order local auto-correlations and its applications to Landsat forestry data. | Milos Stojmenovic, Takumi Kobayashi, Nobuyuki Otsu |
| 2010 | ICONIP | Bilinear Formulated Multiple Kernel Learning for Multi-class Classification Problem. | Takumi Kobayashi, Nobuyuki Otsu |
| 2010 | ICONIP | Logistic Label Propagation for Semi-supervised Learning. | Kenji Watanabe, Takumi Kobayashi, Nobuyuki Otsu |
| 2010 | ICPR | Von Mises-Fisher Mean Shift for Clustering on a Hypersphere. | Takumi Kobayashi, Nobuyuki Otsu |
| 2010 | ICPR | Bag of Hierarchical Co-occurrence Features for Image Classification. | Takumi Kobayashi, Nobuyuki Otsu |
| 2010 | Interspeech | Audio-based sports highlight detection by fourier local auto-correlations. | Jiaxing Ye, Takumi Kobayashi, Tetsuya Higuchi |
| 2009 | ICASSP | Color image feature extraction using color index local auto-correlations. | Takumi Kobayashi, Nobuyuki Otsu |
| 2009 | ICIP | Efficient reduction of support vectors in kernel-based methods. | Takumi Kobayashi, Nobuyuki Otsu |
| 2008 | ECCV | Image Feature Extraction Using Gradient Local Auto-Correlations. | Takumi Kobayashi, Nobuyuki Otsu |
| 2008 | ICPR | Cone-restricted subspace methods. | Takumi Kobayashi, Nobuyuki Otsu |
| 2007 | ICIP | Image Matting in the Framework of Quantification IV. | Takumi Kobayashi, Tadaaki Hosaka, Nobuyuki Otsu |
| 2007 | ICONIP | Application of the Unusual Motion Detection Using CHLAC to the Video Surveillance. | Kenji Iwata, Yutaka Satoh, Takumi Kobayashi, Ikushi Yoda, Nobuyuki Otsu |
| 2005 | MVA | A New Scheme for Image Recognition Using Higher-Order Local Autocorrelation and Factor Analysis. | Naoyuki Nomoto, Yusuke Shinohara, Takayoshi Shiraki, Takumi Kobayashi, Nobuyuki Otsu |
| 2004 | ICPR | Action and Simultaneous Multiple-Person Identification Using Cubic Higher-Order Local Auto-Correlation. | Takumi Kobayashi, Nobuyuki Otsu |