About me
Past
I obtained a Masters of Engineering (MEng) in Computer Science (2010) and a Masters of Philosophy (MPhil) in Computer Vision & Machine Learning (2014) both from Imperial College, London. In between these, I worked as a software engineer at Goldman Sachs for 2 years.
My previous research interests were focussed on Random Forests, mostly applied to pose estimation, both for articulated hand pose and for multiple instances of textureless 3D objects. This research resulted in two conference publications (CVPR and ECCV) and two journal papers (PAMI) as well as two filed patents. For more information, see below.
After this I went on to work at a augmented reality and visual search startup called Blippar. In this role, I co-wrote the augmented reality tracking engine used on mobile devices (for 2D planar tracking) as well as worked on applying deep learning to large-scale visual object recognition (Blippar’s visual browser). Our work in this field has been showcased in many major news outlets such as The Daily Mail, The Telegraph and The Wall Street Journal. Additionally, it has been demoed on major TV stations such as Bloomberg CNBC, Fox News NY and BBC London News.
Present
Currently I work as a Research Engineer for Twitter currently focussed on recommender systems with application towards ranking tweet timelines. Previously my research interests are were around using computer vision for lower-level vision tasks such as super resolution and artifact removal, with possible applications to image/video compression and enhancement.
Future
…who knows, get in contact with me at alykhan[dot]tejani[at]gmail[dot]com or @alykhantejani
Open Source
I am a firm believer in the power of open source to accelerate innovation. I am the author of ignite, maintainer of torchvision and contributor to PyTorch (commits).
Publications
D. Guo, S. I. Ktena, F. Huszar, P. K. Myana, W. Shi, A. Tejani
Deep Bayesian Bandits: Exploring in Online Personalized Recommendations
[PDF][Video]
C. Zhang, Y. Liu, Y. Xie, S. I. Ktena, A. Tejani, A. Gupta, P. K. Myana, D. Dilipkumar, S. Paul, I. Ihara, P. Upadhyaya, F. Huszar, W. Shi
Model Size Reduction Using Frequency Based Double Hashing for Recommender Systems
[PDF][Video]
L. Belli, S. I. Ktena, A. Tejani, A. Lung-Yut-Fon, F. Portman, X. Zhu, Y. Xie, A. Gupta, M. Bronstein, A. Delić, G. Sottocornola, W. Anelli, N. Andrade, J. Smith, W. Shi
Privacy-Preserving Recommender Systems Challenge on Twitter’s Home Timeline
[PDF][Project Page]
B. Steiner, Z. DeVito, S. Chintala, S. Gross, A. Paszke, F. Massa, A. Lerer, G. Chanan, Z. Lin, E. Yang, A. Desmaison, A. Tejani, A, Kopf, J. Bradbury, L. Antiga, M, Raison, N, Gimlelshein, S. Chilamkurthy, T. Killeen, L. Fang, J. Bai
PyTorch: An Imperative Style, High-Performance Deep Learning Library
In Advances in Neural Information Processing Systems (NeurIPS), 2019.
[PDF]
S. I. Ktena, A. Tejani, L. Theis, P. K. Myana, D. Dilipkumar, F.Huszar, S. Yoo, W. Shi
Addressing Delayed Feedback for Continuous Training with Neural Networks in CTR prediction
Proc. of ACM Conf. on Recommender Systems (RecSys), 2019.
[PDF]
L. Theis, I. Korshunova, A. Tejani, F. Huszár
Faster gaze prediction with dense networks and Fisher pruning
[PDF]
C. Ledig, L. Theis, F. Huszár, J. Caballero, A. Cunningham, A. Acosta, A. Aitken, A. Tejani, J. Totz, Z. Wang, W. Shi
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
Proc. of IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2016.
[PDF]
A. Tejani, D. Tang, R. Kouskouridas, T-K. Kim
Latent-Class Hough Forests for 3D Object Detection and Pose Estimation
Proc. of European Conference on Computer Vision (ECCV), 2014.
[PDF][Demo on YouTube][Project Page]
D. Tang, H.J. Chang*, A. Tejani*, T-K. Kim
Latent Regression Forest: Structured Estimation of 3D Articulated Hand Posture
Proc. of IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2014.
*indicates equal contribution
[PDF][Demo on YouTube][Project Page]
Journal Papers
A. Tejani*, R. Kouskouridas*, A. Doumanoglou, D. Tang, T-K. Kim
Latent-Class Hough Forests for 6 DoF Object Pose Estimation
Trans. on Pattern Analysis and Machine Intelligence (PAMI), 2018.
*indicates equal contribution
[PDF][Project Page]
D. Tang, H.J. Chang, A. Tejani, T-K. Kim
Latent Regression Forest: Structured Estimation of 3D Hand Poses
Trans. on Pattern Analysis and Machine Intelligence (PAMI), 2016.
[PDF](Draft)
Patents
U.S. [61/831,255]: Estimator Training Method and Pose Estimating Method Using Depth Image
Korea [10-2013-0131658]: Estimator Training Method and Pose Estimating Method Using Depth Image