Human Robot Interaction Course
Human-Robot Interaction (HRI) studies how people and robots communicate, collaborate and coexist safely. Student projects combine perception (vision, audio), decision-making and robot behaviour, often evaluated with user studies or public interaction datasets.
Below: 90+ HRI topics with tools and representative datasets. Suitable for CSE, ECE, Mechatronics and interdisciplinary programmes.
Tools & Frameworks
ROS2
MediaPipe
OpenPose
PyTorch
Hugging Face
Gazebo
Best Human-Robot Interaction Project Topics (90+)
Topics with tools and datasets / sources.
| # | HRI Project Topic | Tools | Datasets / Sources |
|---|---|---|---|
| Gesture Recognition & Body Language | |||
| 01 | GestureStatic Hand Gesture Recognition for Robot Commands | MediaPipe · OpenCV · scikit-learn | JESTER · custom gestures |
| 02 | GestureDynamic Hand Gesture Sequence Classification for Mobile Robots | LSTM · MediaPipe · ROS2 | JESTER · NVIDIA Dynamic Hand |
| 03 | GestureFull-Body Pose Estimation for Intent Prediction (Approach / Leave) | OpenPose · MediaPipe · PyTorch | NTU RGB+D · custom lab |
| 04 | GesturePointing Gesture Detection for Shared Attention with a Robot | MediaPipe · geometry · ROS2 | custom pointing videos |
| 05 | GestureSign Language Recognition Pipeline for Assistive HRI | MediaPipe · Transformers · Python | WLASL · ASL datasets |
| 06 | GestureMulti-User Gesture Disambiguation in Group Interaction | tracking · clustering · ROS2 | multi-person pose sets |
| 07 | GestureGesture-Controlled Arm Manipulator in Simulation | MoveIt · MediaPipe · Gazebo | synthetic + lab gestures |
| 08 | GestureRobust Gesture Recognition under Occlusion and Clutter | augmentation · ensemble · OpenCV | JESTER · noisy variants |
| 09 | GestureContinuous Gesture Spotting vs Isolated Classification | HMM / CTC · deep models | gesture continuous corpora |
| 10 | GestureCross-Cultural Gesture Mapping for Social Robot Behaviours | survey + ML · Python | literature / small user study |
| Speech, Dialogue & Natural Language HRI | |||
| 11 | SpeechSpeech Command Recognition for Robot Navigation Control | SpeechCommands · TF/PyTorch · ROS2 | Google Speech Commands |
| 12 | SpeechTask-Oriented Dialogue System for Service Robot Scenarios | Rasa / HF · dialogue state | MultiWOZ-style · custom |
| 13 | SpeechWake-Word and Keyword Spotting on Edge for Always-On HRI | TFLite · edge models | Hey-Snips / custom wake |
| 14 | SpeechMultilingual Voice Interface for Public-Space Robots | Whisper / HF · ROS2 | Common Voice · multilingual |
| 15 | SpeechDialogue Repair and Clarification Strategies when ASR Fails | dialogue manager · Python | Wizard-of-Oz logs |
| 16 | SpeechProsody and Emotion from Speech for Adaptive Robot Responses | librosa · SER models · PyTorch | IEMOCAP · RAVDESS |
| 17 | SpeechSituated Language Understanding: Referring to Objects in Scene | vision-language · ROS2 | RefCOCO-style · lab scenes |
| 18 | SpeechLLM-Powered Conversational Companion with Safety Filters | HF Transformers · guardrails | dialogue safety datasets |
| 19 | SpeechTurn-Taking and Interruption Handling in Human-Robot Dialogue | VAD · timing policies · ROS2 | conversation timing logs |
| 20 | SpeechVoice-Based Authentication / Speaker ID for Personalised HRI | speaker embeddings · Python | VoxCeleb concepts |
| Eye Gaze, Attention & Engagement | |||
| 21 | GazeGaze Estimation for Joint Attention with a Social Robot | OpenCV · MediaPipe FaceMesh | MPIIGaze · Gaze360 concepts |
| 22 | GazeEngagement Detection from Head Pose and Eye Contact | MediaPipe · classifiers | engagement labelled videos |
| 23 | GazeMutual Gaze Detection and Contingent Robot Behaviour | face mesh · ROS2 · policy | lab interaction recordings |
| 24 | GazeAttention-Aware Information Presentation by a Guide Robot | gaze heatmaps · dialogue | museum / guide scenarios |
| 25 | GazeDistraction Detection during Collaborative Tasks | head pose · activity features | collaborative task videos |
| 26 | GazeChild vs Adult Gaze Patterns in Child-Robot Interaction | age-aware models · ethics | child-robot interaction sets |
| 27 | GazeScreen vs Robot Gaze Preference Studies with Simple Metrics | eye-tracking analysis · R/Python | lab eye-tracking study |
| 28 | GazeGaze-Contingent Robot Motion for Smooth Shared Attention | control policies · ROS2 | simulation + user study |
| Emotion, Affect & Empathy | |||
| 29 | EmotionFacial Emotion Recognition for Adaptive Social Robot Responses | FER models · AffectNet · PyTorch | AffectNet · FER2013 |
| 30 | EmotionMultimodal Emotion Recognition (Face + Voice + Posture) | fusion networks · PyTorch | IEMOCAP · custom multimodal |
| 31 | EmotionStress / Cognitive Load Detection for Adaptive Task Allocation | physiological proxies · ML | WESAD concepts · lab |
| 32 | EmotionEmpathic Response Generation for Consoling / Support Scenarios | NLG · HF · templates | empathy dialogue corpora |
| 33 | EmotionEmotion Contagion Modelling: Robot Affect Display Impact | affective displays · user ratings | user study design |
| 34 | EmotionValence-Arousal Continuous Affect Estimation for HRI | regression models · face/voice | SEWA · AffectNet VA |
| 35 | EmotionCross-Cultural Emotion Display Preferences for Social Robots | survey + perception study | cross-cultural ratings |
| 36 | EmotionPrivacy-Preserving Affect Sensing on Edge Devices | on-device FER · TFLite | local inference only |
| Collaborative Robots (Cobots) & Shared Tasks | |||
| 37 | CobotHand-Over Timing and Trajectory for Safe Object Exchange | MoveIt · ROS2 · timing policies | handover datasets / lab |
| 38 | CobotForce / Torque Based Intention Estimation during Collaboration | FT sensors · ML · ROS2 | collaborative assembly logs |
| 39 | CobotShared Workspace Collision Avoidance with Human Prediction | skeleton tracking · planning | NTU · lab motion capture |
| 40 | CobotAssembly Task Allocation between Human and Cobot | optimisation · ROS2 | assembly sequence data |
| 41 | CobotLearning from Demonstration for Collaborative Manipulation | LfD · dynamic movement primitives | demonstration trajectories |
| 42 | CobotLegible Motion Generation so Humans Predict Robot Intent | trajectory optimisation · ROS2 | user prediction studies |
| 43 | CobotErgonomics-Aware Cobot Assistance (Reach / Load Reduction) | biomechanics proxies · planning | posture estimation data |
| 44 | CobotFailure Recovery Strategies when Human Interrupts Cobot Task | state machines · ROS2 | interruption scenarios |
| 45 | CobotMulti-Step Collaborative Assembly with Verbal + Gesture Cues | dialogue + pose · ROS2 | task protocols |
| 46 | CobotDigital Twin of Human-Cobot Cell for Offline Programming | simulation · Gazebo / Isaac concepts | cell CAD + trajectories |
| Social Robotics & Proxemics | |||
| 47 | Proxemics: Comfortable Approach Distances for Service Robots | distance policies · user study | proxemics literature + lab |
| 48 | Social Navigation: Passing and Queuing in Crowded Spaces | social force / RL · ROS2 Nav2 | crowd simulation · THÖR concepts |
| 49 | Greeting and Farewell Behaviours for Reception Robots | behaviour trees · speech · face | scripted interaction logs |
| 50 | Personality Trait Expression through Robot Motion and Speech | parameterised behaviours · ratings | Big Five rating studies |
| 51 | Group Interaction: Addressing the Right Person in a Crowd | multi-person tracking · attention | group interaction videos |
| 52 | Storytelling / Education Robot with Adaptive Pacing | dialogue · engagement sensors | education interaction logs |
| 53 | Pet-like Companion Behaviours and Attachment Metrics | behaviour design · surveys | companion robot studies |
| 54 | Cultural Norms in Greeting Gestures and Eye Contact | cross-cultural design · evaluation | cultural HRI literature |
| 55 | Robot Humour and Appropriateness in Workplace Settings | NLG · user appropriateness ratings | humour corpora · ratings |
| 56 | Long-Term Interaction: Habituation and Engagement Decay | longitudinal metrics · logs | multi-week pilot data |
| Multimodal Fusion & Perception | |||
| 57 | MultiEarly vs Late Fusion of Vision and Audio for Intent Recognition | PyTorch · multimodal nets | custom / Ego4D concepts |
| 58 | MultiAudiovisual Speech Recognition in Noisy Robot Environments | AV-ASR · HF · enhancement | LRS / noisy speech sets |
| 59 | MultiSensor Selection Policies under Bandwidth / Power Limits | bandits · edge sensing | multi-sensor logs |
| 60 | MultiSynchronisation of Gesture and Speech for Natural Behaviour | timeline alignment · ROS2 | behaviour scripting tools |
| 61 | MultiEgocentric Vision for Wearable-Assisted HRI | ego models · OpenCV | Ego4D concepts |
| 62 | MultiContact and Non-Contact Sensing Fusion for Safe Approach | RGB-D · proximity · FT | lab multi-sensor |
| 63 | MultiContext Recognition (Home / Office / Factory) for Mode Switching | scene classification · ROS2 | Places / indoor scenes |
| 64 | MultiUncertainty-Aware Multimodal Decision Making for HRI | Bayesian / ensembles · policies | calibrated predictions |
| Safety, Trust, Ethics & Evaluation | |||
| 65 | SafetySafety Zones and Speed Reduction near Detected Humans | skeleton · speed scaling · ROS2 | ISO/TS concepts · lab |
| 66 | SafetyEmergency Stop and Compliant Behaviour under Unexpected Contact | control modes · FT · ROS2 | contact event logs |
| 67 | SafetyTrust Measurement Scales and Behavioural Proxies in HRI | surveys · interaction metrics | trust questionnaires |
| 68 | SafetyExplainable Robot Decisions for User Trust Calibration | XAI · verbal explanations | decision log + ratings |
| 69 | SafetyPrivacy-by-Design: On-Device Processing vs Cloud for HRI Data | architecture comparison · ethics | policy analysis |
| 70 | SafetyBias and Fairness in Perception Models used for HRI | Fairlearn · subgroup metrics | demographic-tagged media |
| 71 | SafetyUser Study Design: Within vs Between Subjects for HRI Metrics | experiment design · statistics | pilot study data |
| 72 | SafetyFailure Transparency: How Robots Should Communicate Errors | dialogue strategies · ratings | error scenario studies |
| 73 | SafetyChildren and Vulnerable Users: Design Constraints for HRI | guidelines · ethics review | literature + expert input |
| 74 | SafetyCybersecurity Basics for Networked Social Robots | threat model · ROS2 security | security checklist |
| Teleoperation, Interfaces & Immersive HRI | |||
| 75 | TeleopShared Autonomy: Blending User Input with Robot Assistance | arbitration policies · ROS2 | teleop trajectory logs |
| 76 | TeleopVR Teleoperation Interface with Force / Visual Feedback | Unity / ROS bridge · haptics concepts | teleop user study |
| 77 | TeleopGesture vs Joystick vs Voice for Mobile Robot Teleoperation | comparative user study · ROS2 | task completion metrics |
| 78 | TeleopLatency Compensation and Predictive Displays for Remote HRI | prediction · UI design | delayed teleop trials |
| 79 | TeleopAR Overlay of Robot Intent on Shared Workspace | AR markers · ROS2 · visualisation | collaborative AR study |
| Applications & Domain-Specific HRI | |||
| 80 | Hospital Guide Robot: Wayfinding Dialogue and Navigation | Nav2 · dialogue · ROS2 | indoor maps · dialogue scripts |
| 81 | Elderly Care Companion: Medication Reminders and Social Chat | dialogue · scheduling · sensors | care scenario protocols |
| 82 | AppClassroom Assistant Robot for Attendance and Simple Q&A | vision · dialogue · ethics | classroom pilot design |
| 83 | AppFactory Floor Gesture Interface for Conveyor / Cell Control | MediaPipe · industrial ROS | factory gesture set |
| 84 | AppMuseum Guide with Multilingual Q&A and Exhibit Detection | vision-language · speech | exhibit images · FAQs |
| 85 | AppTherapy Support Robot: Emotion Tracking and Activity Prompts | FER · activity plans · ethics | therapy session design |
| Advanced & Capstone | |||
| 86 | AdvFoundation Models for HRI: Vision-Language Action Prototypes | HF · VLM · robot API | lab task suite |
| 87 | AdvSim-to-Real Transfer of Social Navigation Policies | Gazebo · RL · domain randomisation | crowd sim → lab |
| 88 | AdvPersonalisation: Adapting Robot Behaviour to User Preferences | preference learning · bandits | multi-session user data |
| 89 | AdvBenchmarking Open-Source HRI Stacks on Common Tasks | ROS2 · standardised tasks | public HRI benchmarks |
| 90 | AdvLongitudinal Field Deployment Metrics for Service Robots | logging · dashboards · stats | multi-week deployment |
| 91 | AdvMultimodal Dataset Collection Pipeline for Local HRI Research | ROS2 bags · annotation tools | custom corpus |
| 92 | AdvCapstone: Full HRI Demo — Sense (vision/audio), Decide, Act, Explain | ROS2 · MediaPipe · dialogue · Nav2 | end-to-end user study |
Datasets are representative public or commonly cited sources. Always check licences and ethics requirements for human-subject data. Contact us for topic refinement, implementation, user-study design, report, PPT and viva support.
Why Choose Us for HRI Projects?
Gesture & Vision
MediaPipe, OpenPose and robust recognition pipelines for robot command and intent.
Speech & Dialogue
Task-oriented dialogue, wake words and LLM-assisted conversation with safety filters.
Cobots & Social
Handovers, shared workspaces, proxemics and social navigation in ROS2.
Trust & Safety
Explainability, fairness, emergency behaviours and user-study evaluation methods.
FAQ — Human-Robot Interaction Projects
Strong topics include gesture-controlled robots, task-oriented dialogue, gaze-based joint attention, multimodal emotion recognition, cobot handovers, social navigation, trust measurement and full multimodal HRI demos with user evaluation.
ROS2, MediaPipe, OpenPose, PyTorch, Hugging Face, Gazebo, OpenCV; datasets such as JESTER, NTU RGB+D, AffectNet, Speech Commands, EuRoC-style interaction logs and custom lab recordings.
Many topics can be completed in simulation (Gazebo) plus webcam/mic perception. Hardware (TurtleBot-class, cobot arms, social robot platforms) strengthens demos when available.
Yes — reference material, perception/dialogue stacks, evaluation notes, university-format report, PPT and viva Q&A, including guidance on ethical user studies where needed.
HRI Project Lab — Bangalore
Perception, dialogue and collaboration support for human-robot interaction projects.
MediaPipe
Gesture Lab
Hand / PoseGesture Lab
Dialogue
& Speech
Task-Oriented& Speech
Gaze &
Engagement
AttentionEngagement
Emotion
Recognition
Affective HRIRecognition
Cobot
Collaboration
Handover / SafetyCollaboration
Social
Navigation
ProxemicsNavigation
Multimodal
Fusion
Vision + AudioFusion
Viva & Report
Support
DocumentationSupport