Skip to content

Commit 6a2506b

Browse files
sohams25sohams25
committed
feat: enrich experience and projects from CV, add Clutterbot logo and personal interests
1 parent 7374a9e commit 6a2506b

2 files changed

Lines changed: 19 additions & 14 deletions

File tree

‎images/ClutterBot_Logo.png‎

6.33 KB
Loading

‎index.html‎

Lines changed: 19 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -104,12 +104,12 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
104104
<p>
105105
During my undergraduate years, I contributed to various student-led projects at BITS Pilani, Goa, including
106106
<a href="https://kratos-the-rover.github.io" target="_blank">Project Kratos</a>,
107-
where I was involved in the Autonomous Subsystem of a Mars Rover prototype. I also served as the Events, Initiatives, and Sponsorship Head at the
107+
where I was involved in the Autonomous Subsystem of a Mars Rover prototype. I also served as the Events and Initiatives Head at the
108108
<a href="https://bpgc-cte.org/" target="_blank">Center for Technical Education, BITS Goa</a>.
109109
</p>
110-
110+
111111
<p>
112-
During my free time, I enjoy playing my ukulele and listening to music. =)
112+
Outside of work, you're usually found brewing Blue Tokai in a French press, bouldering, or planning for treks in the Himalayas.
113113
</p>
114114

115115
<!-- <p style="text-align: center; font-size: medium;">
@@ -161,7 +161,7 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
161161
<tr>
162162
<td style="padding:10px;width:25%;vertical-align:middle">
163163
<div class="one">
164-
<img src='images/Clutterbot.png' width="120" onerror="this.style.display='none'">
164+
<img src='images/ClutterBot_Logo.png' width="120">
165165
</div>
166166
</td>
167167
<td style="padding:10px;width:75%;vertical-align:top">
@@ -171,7 +171,7 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
171171
<br>
172172
<br>
173173
<p>
174-
Working on applied machine learning for home robotics, building perception and learning-based components for autonomous systems.
174+
Built an end-to-end pipeline for training and deploying segmentation and detection models for indoor navigation, packaging deployment deliverables for the DAM team. Migrated a complex perception system to the Qualcomm DragonWing QCS6490P chipset and NVIDIA IoT kits, improving edge inference speed by ~30% in barebones Linux environments using GStreamer. Also designed and deployed a custom 6DOF pose estimation proof-of-concept pipeline to ensure reliable target locking and rebasing during the hardware migration.
175175
</p>
176176
</td>
177177
</tr>
@@ -190,7 +190,7 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
190190
<br>
191191
<br>
192192
<p>
193-
Developed gaze estimation models using Computer Vision and Deep Learning for Affective Computing in education. Analyzed correlations between gaze patterns and student engagement to establish quantitative measures of attention.
193+
Developed and benchmarked a Vision Transformer (ViViT) for classroom activity recognition on a 927-clip EduNet subset, achieving 88% test accuracy by fine-tuning a Kinetics-400 pre-trained model with a 224x224 video preprocessing pipeline. Validated generalizability on an independent 100-video dataset (72% accuracy) and generated gradient-based saliency maps to confirm focus on key regions (e.g. raised hand, board writing). Analyzed gaze and activity patterns to derive quantitative student engagement measures.
194194
</p>
195195
</td>
196196
</tr>
@@ -210,7 +210,7 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
210210
<br>
211211
<br>
212212
<p>
213-
Optimized a SAM-based image segmentation GUI, enabling efficient AI-assisted annotation, and developed an automated mechanism to aggregate segmented regions into sub-scenes, enhancing the semantic understanding of complex images.
213+
Implemented and optimized a PyQt5-based GUI integrating SAM for image segmentation, enabling efficient AI-assisted annotation. Developed and automated a mechanism to aggregate individually segmented regions into larger sub-scenes using advanced image processing techniques, advancing the semantic understanding of complex images and reducing manual effort.
214214
</p>
215215
</td>
216216
</tr>
@@ -229,7 +229,7 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
229229
<br>
230230
<br>
231231
<p>
232-
Worked under under the supervision of <a href="https://www.ceeri.res.in/profiles/dhiraj-sangwan/" target="_blank">Dr Dhiraj Sangwan</a> on developing a comprehensive image restoration pipeline, enhancing my skills in Computer Vision and Deep Learning, covering Image Segmentation, Image Processing, and GANs.
232+
Worked under the supervision of <a href="https://www.ceeri.res.in/profiles/dhiraj-sangwan/" target="_blank">Dr. Dhiraj Sangwan</a> on restoring and segmenting deteriorated Rajasthani murals using deep learning models like U-Net++, DeepLabV3+, PSPNet, and FPN. Developed a synthetic damaged-image generation pipeline using GANs (StyleGAN2-ADA) and crafted diverse binary masks to simulate complex real-world mural damage. Implemented a state-of-the-art inpainting pipeline achieving an SSIM score of 0.9812 for reinstating missing sections.
233233
</p>
234234
</td>
235235
</tr>
@@ -318,11 +318,13 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
318318
<td style="padding:20px;width:25%;vertical-align:top">
319319
<papertitle><big>CloSe++</big></papertitle>
320320
<br>
321-
<span style="color:gray">2024</span>
321+
<span style="color:gray">Feb '24 - May '24</span>
322+
<br>
323+
<span style="color:gray">Supervised by Dr. Garvita Tiwari</span>
322324
</td>
323325
<td style="padding:20px;width:75%;vertical-align:top">
324326
<p>
325-
Extension of the <a href="https://virtualhumans.mpi-inf.mpg.de/close3dv24/" target="_blank">CloSe</a> framework for 3D human cloth segmentation on point clouds, exploring improved supervised approaches for fine-grained garment label prediction.
327+
Extended the <a href="https://virtualhumans.mpi-inf.mpg.de/close3dv24/" target="_blank">CloSe</a>-Net framework for fine-grained 3D clothing segmentation from coloured point clouds. Refined edge detection mechanisms to produce sharper boundaries between clothing types, addressing the blurriness in the original model, and automated clothing-type detection to remove manual input and improve usability.
326328
</p>
327329
</td>
328330
</tr>
@@ -331,11 +333,13 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
331333
<td style="padding:20px;width:25%;vertical-align:top">
332334
<papertitle><big>Project Visio</big></papertitle>
333335
<br>
334-
<span style="color:gray">2022 - 2023</span>
336+
<span style="color:gray">Oct '22 - Aug '23</span>
337+
<br>
338+
<span style="color:gray">Supervised by Prof. Sougata Sen</span>
335339
</td>
336340
<td style="padding:20px;width:75%;vertical-align:top">
337341
<p>
338-
Student-led computer vision initiative at BITS Pilani, Goa focused on building deep-learning-based perception modules and exploring applied vision research problems.
342+
Engineered smart-glasses to aid the visually impaired, integrating computer vision techniques for object detection and scene understanding. Developed Tiny-ML models for efficient inference on ESP microcontrollers and designed a companion Android app to interface with the smart-glass system, ensuring seamless real-time assistance over WiFi.
339343
</p>
340344
</td>
341345
</tr>
@@ -344,11 +348,11 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
344348
<td style="padding:20px;width:25%;vertical-align:top">
345349
<papertitle><big><a href="https://kratos-the-rover.github.io" target="_blank">Project Kratos</a></big></papertitle>
346350
<br>
347-
<span style="color:gray">2022 - 2023</span>
351+
<span style="color:gray">Sept '22 - Jan '23</span>
348352
</td>
349353
<td style="padding:20px;width:75%;vertical-align:top">
350354
<p>
351-
Contributed to the Autonomous Subsystem of a Mars Rover prototype as part of BITS Goa's student-led rover team, working on perception and navigation for autonomous traversal tasks.
355+
Contributed to the Autonomous Subsystem of BITS Goa's student-managed Mars Rover. Optimized autonomous navigation in ROS, composed and modified path-planning algorithms (A*, Dijkstra, SLAM) in the Gazebo simulator interfaced with an Nvidia Jetson Xavier, and improved navigation through OpenCV-based object detection.
352356
</p>
353357
</td>
354358
</tr>
@@ -387,6 +391,7 @@ <h1 style="text-align:center"><name>Soham Shinde</name></h1>
387391
<img src="images/Central_Electronics_Engineering_Research_Institute_Logo.png" width="92" style="padding: 3px;">
388392
<img src="images/TCSR_Logo.png" width="92" style="padding: 3px;">
389393
<img src="images/NTU.png" width="92" style="padding: 3px;">
394+
<img src="images/ClutterBot_Logo.png" width="92" style="padding: 3px;">
390395
</p>
391396
<p style="text-align: center;">This template is a modification to Jon Barron's website</a>. Find the source code to my website <a href="https://github.com/sohams25/sohams.github.io" target="_blank">here</a>.</p>
392397
</td>

0 commit comments

Comments
 (0)