Positions

If you are an Imperial College UG student, you can get involved with us through:

  •   Final-Year Projects. Every year the group offers final year projects related to our group’s research.
  •   Industrial Placements: 3rd Year students can take their placement in our group.
  •   UROP: We offer summer placements to students of all years.  

 

Entry Requirements

Strong academic record in Digital System Design, Programming, Mathematics. Good knowledge of Machine Learning and image processing is desirable.

FPL 2024

May 2024
Our paper on Hardware-aware Sparsity Search (HASS) is accepted to FPL!

FCCM 2024

May 2024
Our team member Petros has presented our work "SMOF" in Orlando. Check the presentation now!

FCCM 2024

Mar 2024
Our paper "SMOF: Streaming Modern CNNs on FPGAs with Smart Off-Chip" is accepted to FCMM. Congrats!

AMD visit

Mar 2024
Our team presented our latest work on the acceleration of AI models on FPGAs in AMD San Jose.

SpatialML workshop

Mar 2024
Our recent work on fpgaConvNet was presented in the SpatialML workshop co-located with ISFPGA’24

cornell tech visit

Feb 2024
Our team member Zhewen Yu shared his work on Accelerator Design Space Exploration in Cornell Tech

Dr Ben Chua

Jan 2024
Congratulations on our team member Ben Chua for passing his PhD viva!

DATE 2024

Nov 2023
Our work on Hardware accelerated Gaussian Belief Propagation in SLAM problems is accepted to DATE

DATE 2024

Nov 2023
Our work, Automate Weights Streaming in Layer-wise Pipelined DNN Accelerators, accepted to DATE2024

FPT 2023

Oct 2023
Our work SATAY: Streaming Architecture Toolflow for Accelerating YOLO Models is accepted to FPT2023

ICCV 2023

Jul 2023
Our work on Window-Based Early-Exit Cascades for Uncertainty Estimation is accepted to ICCV2023

FPL 2023

May 2023
Our paper on Exploiting CNN Post-Activation Sparsity in Streaming Architectures is accepted to FPL

FPL 2023

May 2023
Our paper on Efficient Accelerator with Layer-Specific Tensor Decomposition is accepted to FPL

FPL 2023

May 2023
Our paper on a toolflow for throughput oriented HAR acceleration on FPGAs is accepted to FPL

ASAP 2023

May 2023
Our work on optimization and mapping of X3D HAR model on FPGAs is accepted to ASAP

FCCM 2023

Apr 2023
Our work on a latency-oriented 3D-CNN accelerator toolflow will show in this year's FCCM

Best Paper Nomination

Apr 2023
Our work on designing efficient FPGA-based accelerators for Early-Exit CNNs accepted to FCCM23

Best Paper Award

Dec 2022
Our work on Selective Classification with OOD data received ACCV Saburo Tsuji Best Paper Award!

ACCV 2022

Sep 2022
Our work on Selective Classification with OOD Data will show in this year's ACCV

ECCV 2022

Sep 2022
Our work On the Usefulness of Deep Ensemble Diversity for OOD Detection is accepted at ECCV workshop

WACV 2023

Sep 2022
Our paper SVD-NAS Coupling Low-Rank Approximation and Neural Architecture Search is accepted to WACV

NANDA 2022

Sep 2022
Prof. Bouganis gave an invited talk on CNN acceleration in the NANDA workshop

RCML 2022

Sep 2022
Prof. Bouganis co-organised the 6th Workshop on Reconfigurable Computing for ML (RCML), at FPL 2022

ASP-DAC 2021

DEC 2020
Our paper on Differential Encoding of Featuremaps for CNN Accelerators has been accepted to ASP-DAC.

ARC 2020

OCT 2020
Our paper on "A Parameterisable FPGA-Tailored Architecture for YOLOv3-Tiny" has been accepted to ARC

EMDL Keynote

SEP 2020
Dr. Bouganis was invited to give a keynote in EMDL 2020 workshop, co-located with Mobicom 2020.

Next Platform Interview

JUL 2020
Watch Dr. Stylianos Venieris' interview on "The next platform" discussing our collaborative work.

SAIC Talk

JUL 2020
Dr. Bouganis was invited to give a presentation of our recent work to Samsung AI Center, Cambridge.

ICML 2020

JUN 2020
Our paper on Multi-Precision Policy Enforced Training (MuPPET) has been accepted to ICML2020.

Dr. Nur Ahmadi

JUN 2020
Congratulations to our group member Nur Ahmadi for passing his PhD Viva

FPL 2020

MAY 2020
Our paper on Caffe Barista: Brewing Caffe with FPGAs in the Training Loop has been accepted to FPL20

Dr. Manolis Vasileiadis

MAY 2020
Congratulations to our group member Manolis Vasileiadis for passing his PhD Viva

PhD studentship

MAR 2020
Exciting opportunity to join our group and work closely with Arm, on probabilistic ML acceleration.

HiPEAC WRC 2020

JAN 2020
Dr. Bouganis co-organised the 14th HiPEAC Workshop on Reconfigurable Computing.

IEEE RAS Conf. 2020

JAN 2020
Our IROS19 work on UAV-based vehicle detection awarded the 3rd place in the IEEE UK&I RAS Conference

DATE 2020

NOV 2019
Our paper on "A Throughput-Latency Co-Optimised Cascade of CNN Classifiers" is accepted at DATE2020.

FPT 2019

NOV 2019
Our paper on “Power-Aware FPGA Mapping of CNNs for On-Device Inference” is accepted at FPT2019.

BMVC 2019

OCT 2019
Our paper on 3D-CNNs for Human Pose Estimation on low-power devices is accepted at BMVC 2019 (pdf).

Welcome Alexander

OCT 2019
We are pleased to welcome Alexander Montgomerie as a PhD Student in iDSL.

RCML19

SEPT 2019
Dr. Bouganis co-organised the 3rd Workshop on Reconfigurable Computing for ML (RCML), in FPL 2019.

ISVLSI 2019 Slides

JULY 2019
Dr. Bouganis presented an overview of our recent work at ISVLSI19 (slides)

Dr. Konstantinos Boikos

JUL 2019
Congratulations to our group member Konstantinos Boikos for passing his PhD Viva

Exhibition Road Festival 2019

JUNE 2019
iDSL participates on the Exhibition Road Festival. Come and see our drones in action in the Arena.

IROS 2019

JUNE 2019
Our paper on Informed Region Selection for UAV-based Vehicle Detection is accepted at IROS2019.

ISVLSI 2019

JUNE 2019
Our invited paper on Efficient Deployment of DNNs on Autonomous Systems will be presented in ISVLSI.

SpatialML Centre

MAY 2019
Our International Centre for Spatial Computational Learning has been announced.

FPL 2019

MAR 2019
Dr.Bouganis with be co-chairing the Program-Committee of FPL 2019, held in Barcelona.

ARC 2019

FEB 2019
Our paper on "A Scalable FPGA-based Architecture for Depth Estimation in SLAM" got accepted in ARC19

Welcome Mario

FEB 2019
We are pleased to welcome Mário Ferreira as a Research Assistant in iDSL.

Future of Drones Workshop

OCT 2018
Dr.Bouganis presented our work in the Future of Drones & UAVs workshop at Austrian Trade Commission

IROS 2018 Talk

OCT 2018
Watch our IROS2018 spolight talk on Self-Supervised CNN-based Autonomous UAV Navigation.

Dr. Stylianos I. Venieris

OCT 2018
Congratulations to our group member Stylianos Venieris for passing his PhD Viva

NG-HPC 2018 Presentation

SEPT 2018
Dr. Bouganis presented our work on fpgaConvNet toolflow in NG-HPC 2018 (slides)

RCML18 Presentation

SEPT 2018
Dr. Bouganis presented our work "Deploying Deep Neural Networks on the Embedded Space" in RCML18

Job Opening

AUG 2018
iDSL is recruiting for a new RA position in the area of Machine Learning and FPGA co-design (click)

RCML'18

JULY 2018
Call for papers: 2nd Workshop on Reconfigurable Computing for Machine Learning in FPL 2018

IROS 2018

JUNE 2018
Our paper on Self-Supervised CNN-based Autonomous UAV Navigation has been accepted at IROS 2018.

TNNLS 2018

JUNE 2018
Our TNNLS paper on fpgaConvNet is now available online (link).

MobiSys18 Workshop

JUNE 2018
We are happy to present our work on Deploying Deep Learning Models in Embedded Devices, at EMDL18.

CNN-to-FPGA Benchmark

MAY 2018
Our CNN-to-FPGA Benchmark Suite, introduced in our CSUR paper, can be found in our webpage (Click)

FPL 2018

MAY 2018
Our paper "CascadeCNN: Pushing the Performance Limits of Quantisation in CNNs" got accepted at FPL18

FPL 2018

MAY 2018
Our paper "f-CNNx: A Toolflow for Mapping Multiple CNNs on FPGAs" has been accepted at FPL 2018.

ARC 2018

APR 2018
Our ARC 2018 paper on FPGA-based LSTMs was nominated for best paper award. (Read pre-print)

Imperial Festival 2018

APR 2018
Our drones were flying in the Robotics Zone of this year's Imperial Festival.

Dr.Bouganis Promotion

APR 2018
We are delighted to announce that our director, Dr.Bouganis, has been promoted to a Reader.

CSUR 2018

MAR 2018 (In Press)
CNN-to-FPGA toolflows Survey and Future Directions paper accepted by ACM Computing Surveys

SysML 2018

JAN 2018
Our latest work on Computational ML is accepted to be presented in the inaugural SysML conference.

WRC'18 Talk

JAN 2018
Dr.Bouganis gave a talk on "Toolflows for mapping CNNs into FPGAs" in HiPEAC WRC 2018. (slides)

WRC at HiPEAC 2018

JAN 2018
Dr.Bouganis will be co-organising the 12th Workshop on Reconfigurable Computing, at HiPEAC 2018.

ARC 2018

JAN 2018
New paper: "Approximate FPGA-based LSTMs under Computation Time Constraints” on ARC2018.

NIPS 2017

NOV 2017
fpgaConvNet will be presented on NIPS2017 workshop on "ML on the Phone and Other Consumer Devices"

DATE 2018

NOV 2017
Special Session on Next Generation Processors and Architectures for Deep Learning has been accepted

RCDL'17

SEPT 2017
1st FPL Workshop on Reconfigurable Computing for Deep Learning co-organised by Dr. Bouganis

IROS 2017

SEPT 2017
A very successful and machine learning oriented IROS conference came to an end.

Dr. Shuanglong Liu

JULY 2017
Congratulations to our group member Shuanglong Liu for passing his PhD Viva

FPL 2017

JUNE 2017
Two papers from our group are accepted on FPL 2017.

Dr. Michail Vavouras

JULY 2017
Congratulations to our group member Michail Vavouras for passing his PhD Viva

UROP Students 2017

JUNE 2017
Our group is excited to welcome 2 UROP students for this summer.